ELF>@e @8 @ * *000 Up$ 4 4  $$ @@   Std @@PtdX X X \ \ QtdRtd 55GNU1:<#.P.xIH}HH: HtH53 %3 @hfhfhfhfhfhfhfhrfhbfh Rfh Bfh 2fh "fh fhfhfhfhfhfhfhfhfhfhrfhbfhRfhBfh2fh"fhfhfhfh fh!fh"fh#fh$fh%fh&fh'rfh(bfh)Rfh*Bfh+2fh,"fh-fh.fh/fh0fh1fh2fh3fh4fh5fh6fh7rfh8bfh9Rfh:Bfh;2fh<"fh=fh>fh?fh@fhAfhBfhCfhDfhEfhFfhGrfhHbfhIRfhJBfhK2fhL"fhMfhNfhOfhPfhQfhRfhSfhTfhUfhVfhWrfhXbfhYRfhZBfh[2fh\"fh]fh^fh_fh`fhafhbfhcfhdfhefhffhgrfhhbfhiRfhjBfhk2fhl"fhmfhnfhofhpfhqfhrfhsfhtfhufhvfhwrfhxbfhyRfhzBfh{2fh|"fh}fh~fhfhfhfhfhfhfhfhfhrfhbfhRfhBfh2fh"fhfhfhfhfhfhfhfhfhfhfhrfhbfhRfhBfh2fh"fhfhfhfhfhfhfhfhfhfhfhrfhbfhRfhBfh2fh"fhfhfhfhfhfhfhfhfhfhfhrfhbfhRfhBfh2fh"fhfhfhfhfhfhfhfhfhfhfhrfhbfhRfhBfh2fh"fhfhfhfhfhfhf%& fD%& fD%& fD%& fD%& fD%~& fD%v& fD%n& fD%f& fD%^& fD%V& fD%N& fD%F& fD%>& fD%6& fD%.& fD%&& fD%& fD%& fD%& fD%& fD%% fD%% fD%% fD%% fD%% fD%% fD%% fD%% fD%% fD%% fD%% fD%% fD%% fD%% fD%% fD%% fD%~% fD%v% fD%n% fD%f% fD%^% fD%V% fD%N% fD%F% fD%>% fD%6% fD%.% fD%&% fD%% fD%% fD%% fD%% fD%$ fD%$ fD%$ fD%$ fD%$ fD%$ fD%$ fD%$ fD%$ fD%$ fD%$ fD%$ fD%$ fD%$ fD%$ fD%$ fD%~$ fD%v$ fD%n$ fD%f$ fD%^$ fD%V$ fD%N$ fD%F$ fD%>$ fD%6$ fD%.$ fD%&$ fD%$ fD%$ fD%$ fD%$ fD%# fD%# fD%# fD%# fD%# fD%# fD%# fD%# fD%# fD%# fD%# fD%# fD%# fD%# fD%# fD%# fD%~# fD%v# fD%n# fD%f# fD%^# fD%V# fD%N# fD%F# fD%># fD%6# fD%.# fD%&# fD%# fD%# fD%# fD%# fD%" fD%" fD%" fD%" fD%" fD%" fD%" fD%" fD%" fD%" fD%" fD%" fD%" fD%" fD%" fD%" fD%~" fD%v" fD%n" fD%f" fD%^" fD%V" fD%N" fD%F" fD%>" fD%6" fD%." fD%&" fD%" fD%" fD%" fD%" fD%! fD%! fD%! fD%! fD%! fD%! fD%! fD%! fD%! fD%! fD%! fD%! fD%! fD%! fD%! fD%! fD%~! fD%v! fD%n! fD%f! fD%^! fD%V! fD%N! fD%F! fD%>! fD%6! fD%.! fD%&! fD%! fD%! fD%! fD%! fD% fD% fD% fD% fD% fD% fD% fD% fD% fD% fD% fD% fD% fD% fD% fD% fD%~ fD%v fD%n fD%f fD%^ fD%V fD%N fD%F fD%> fD%6 fD%. fD%& fD% fD% fDU1HAVAUATSHPdH%(HE1HUHUHUHUHUHUOHMHUHxxHuI~H=[[HHu4H H84gH=I+HHH5NHIċxȉuHMQHq I9D$t;H[ H5oH8A$!A$L1LHA A$xA$uL^HA HuH H5H8|LuLeH]=HA v#H5moH H81ylA !H~ H5oH81xGHGA uHN H5pH82H. H5pH8HH I}pH0ǂHԝH=HMHULHu}H@ =w1HuHHHEHyR HEHeIċxȉuHMt11LlA$yA$uLHMHUHuI}x|H}H}H}HH=:L{LsHk1HUdH+%(tHP[A\A]A^]UH=}HSQ6HH }HW HHH5cH |HW HH5מ>H |HW HH5H |HmW HH5H |H@W HH5ςaH Z|HSW HH5t誂<H 5|H6W HH5[腂H |HW HH5B`H {HV HH5);xȉuHH= HHH {HV HH5|H |HV HH5ŁWH HyV HH5蠁2H +|HDV HH5ם{ H |HV HH5VH )|HBV HH51H ,|HU HH5 H '}HU HH5ut}H ~HU HH5Yƀt\H ~HU HH5?襀t;H l~HuU HH5/脀txȉuH11HZ[]UHAWEAVIAUIHATSAQHt>H; Iu1AtLLLA$x1A$u)L~H H8xt1Z[A\A]A^A_]UHAVAUIATSHx?Ht^H Hu H H9tHg H5PH88+L%WV Mu&H5LIHu(L1A$=wA$LHHAxAuLHtHIHt#AH HLHwyIwAH LLH~LxAH oLLHn)xE1H bLLH^ xH[A\A]A^]UHSQ(H)< Hu `HcR =wHOR H; H5 H= HP H H; HtH R HQ 1H5Q H; HyHQ H5Q 1H; HOH  H 1H5 Hc; HHS H5D 1pHA; HH H5Q 1FH; HH5 1#H; HHt H5P 1H: H}H LP 1H 5 H5. IH: HBH  HHP 1H5*P H: HH &P HO 1H5 \H]: HHG H5 12H;: HH5O 1HH: Z[]UHK fHnHAUL-]9 ATLxSH Q~ fl) 9 H;HteC t0@tH3sHc$t1ITSsHcHtHI$HHt7HIW[HN HtןBHN Hu IHN HtHN HtHN Ht1HN HttHuN Ht ^HgN HhDHUN HN2*HCN H4H1N H75*HN H1 H N H11H=ݖ HM HHHHM Z[A\A]]UHATSH@dH%(HE1HHH B HJ L6 HyG H=6 IHHuzH}7 H2HI H I IL|6 H-G H=6 HHHY? HzH.7 HH H HF IL6 H=6 H~WI 09 )yH6 HH-I H I IL5 HF H=5 HHH!E HHA HhyH6 H%H G H1F IL4 H=5 H~ !@  F ~"J )~H @ H ))xH6 HHKH H = IL3 HE H=3 HHHo< HH= HHKH HxxH5 H5HG H ? IL2 H0E H=!3 HHHG HHG HxHK5 HHsG H H IL1 HD H=l2 HHHG HH> HHB HHD HH> HwH4 HAHF H @ IL/ H )~>  J> ~F A |A )))vH+4 HH{B H ,< IL. HC H=L. H~F H@ )vH3 HAHE ILB( HHH: HHF HH1E HHC HH: HHE HHG HHF HH@ HHE HHC HH> HHe? HHE H H)@ H(H> H0He9 H8HC H@H< HHH; HPH : HXHC H`HB HhHs> HpHD HxH: HEHF HEH= HEH@ HEH< HEHp< HEHF H ^: HA HEH@ H=+ HEH8 HEH/F HEH: HEHiC HEYtH1 HHC H E IL$ HA H=% HHH> HH< HHC HHK= HH8 HH7 HHA8 HH3> HH=< HHB HHD HvsH0 H3HB H C ILm# H.@ H=# HHHB HH: HHn@ HH: HrH]0 HH -? H? IL" H=" H~B wB ~ B ) < )rH/ HKHA H B IL! HF? H=" HHHA HHA HH9 HHx? HH: HH8 HqHi/ HHQA H 9 IL H> H= HHHEA HH@ HHQA H~qH. H;H 8 HG> IL H= H~7 A ~ @ ) 7 )qH. HHw@ H 8= IL H= H=  HHHk7 HHU7 HH< HHi@ HpH&. HSH? H 6 IL  HN= H=? HHH6 HH@ H1pH- HH? H B< ILH H< H=z HHHu6 HH; HH? HoH^- H{H? H ? ILu Hv< H= HHH:? HgoH- H$H> H ? IL H< H= HHH5 HH> HoH, HHb> H @ IL H; H=+ HHH: HHx8 HHb> HnHG, HLH= H < IL& HG; H=X HHH2 HH= H*nH+ HH= H C@ ILa H: H= HHH1 HH= HmH+ HH%= H ; IL H}: H= HHHI1 HH3= H`mH0+ HH 07 H): IL H= H~< < ~ < ) <7 )lH* HH Y5 H9 IL H=I H~< {< ~ #< ) 6 )lHr* HOH 6 H[9 ILI H= H~; < ~ ; ) n6 )+lH* H H; H l6 IL H8 H= HHH6 HH; HH; HkH) Hu H; H a: IL Hp8 H= HHH: HH&; HSkHK) H H: H L= IL H 8 H= HHH5 HHY: HH: HjH( H H@: H < IL H7 H=Q HHH4 HH5 HH9 HH2: HH3 HH2 HH2 HH*2 HH<7 HH67 HH`7 HiH( H H]9 H . IL H6 H= HHHA5 HH7 HH]9 HHw0 HHq2 HH2 HH%2 HH7/ HH7 HH7 HH2 HH6 HH4 HhH' H H^8 H 4 IL H5 H= HHHB4 HH6 HH^8 HH1 HHR/ HH\/ HH./ HH5 HH3 HH2 H)hHA& H H7 H 6 IL H4 H= HHH1 HH7 HgH% H H$7 H : IL H|4 H= HHHH+ HH27 H_gH% H H6 H / IL H4 H=( HHH4 HH6 HfH*% HHZ6 H / IL H3 H=c HHH2 HHp2 HH 3 HHL6 HH( HH/ HH3 HH|3 HH~3 HHX0 HHB0 HHD0 H fHA$ HHi5 H Z0 IL H2 H= HHH53 HHw5 HeH# HaH5 IL HHH=0 HH+ HH5 HH* HHe6 HH'0 HH4 HH, HHM7 HH5 HH96 HH#4 HH6 HH+ H Hq* H(Hs. H 0 Hu1 H0H2 H=@ H8H, H@_dH" HH3 IL] HHH/ HHj2 HH3 HHv* HH, HH4 HH0 HH1 HHh1 HH. HH0 HH4 HHx1 HHJ/ H H. H(Hn/ H . H00 H0H, H= H8H2 H@HW. HHH0 HPH4 HXH2 H`H7. HhbH$! HH42 ILR HHH( HH/ HHI2 HHs- HH. HH. HH3 HHK( HH, HH. HH 4 HH1 HHE+ HH1 H H3 H(H ;- H. H= aH HvHQ( H R( IL Hq. H= H~=;1 H@=-+ ~52 )~-+ 5* ~%3 -% ~' %% ~% % ~ % 2 ~2  D* D~k0 0 D,% )D)))))) )0`H HdH?( H X. IL H_- H= H~%I% H%- ~+% )~4% %% ~ % ) ~>1  ) ~-/ 0 -Q2 )))))_HB HH% H c/ IL H, H= H~|$ HN( ~>) )~ / * ~1  ) ~%. $ %1 ))))#_H HH , H+ ILZ H= H~ #  m# ~' )~>. . 1 ))^H HtgH % Hs+ ILa H= H~- & )Y^H Hty1)ȉuHtxȉuH^HUdH+%(tH@[A\]UH=, HY,Hu H=- >,HtH=*! -,H HtH=z& ,H HtH= +H HtH=: +H0 HtH=j +H1 HiH= +H1 HMH= +H H1H=n y+HB1 HH= ]+H1 HH=% A+H1 HH=$ %+HH= +H HH= *H} HH= *HH- ]UHGHuHI HH5O\H81(1Ht&H;W t H HH5U\H81]UHATSHUHdH%(H]HH5' ALeMu S1H޺LÅxLHEdH+%(tMZY[A\]UHAVIAUIATISHÅuULH5* LGt:H5* LE1LL1HHtxȉu P[A\A]A^]HHtHPuHUHAUATSHQHPHtwH HqH9~bHD uHPH H5x[H81vkHu,H t"HHHSH5v[Hg H81E:HyHHA3HAEt kAZD[A\A]]U1HATIHS1HHt3I$H5[( HxxȉuH1 H[A\]UH5( HATSHHtM1H`IHu HuHG H5 [H8xȉtL HH[A\]UHAWAVE1AUATISH(LPHA~Ht HLcHcAHHHIEN4HEL9}KD HIHuII$HL9tH9u/HLEHMHuHMLEHuHHAHL9tHu?IcH57ZID HHI$HPH; H81HHHjH]1H([A\A]A^A_]UHAWIHAVEAUIATISAQ)HHH@u H LLH5YH81qLK(HC Mt I9LLIM9s#H0 MLLH5YH81>-Au2I9s-RL1MPMH~Y11Y^y H1uHeH[A\A]A^A_]UH=rHSQQHAHHHwH5rH HxȉuHH=prHHyA HHpH5Cr{H HJxȉuHH=rHHA HHqH5qHk HxȉuH&H=q:HHA HHvH5qH HAH HHvqH5iqH H_A0HHNqH58q_H H.A`HH'qH5q.H HAHHpH5pHr HAHHpH5pHI HAHHpH5tpH  HjAHH|pH5CpjH H9AHHIpH5p9H HAHH(pH5oH HAHHpH5oH| HAHHoH5oHS HuAHHoH5NouH* HDAHHoH5oDH HAHH^oH5nH HxȉuHH=-o/HHA`HHjH5oH HA@HHnH5nHZ HtXHHtGAHHnH5nGH$ HtxȉuHS11HeZ[]UHHHXH`HhLpLxt )E)M)U)])e)m)u)}dH<%(HHDž0Dž40HHEH8HPH@HtHH=wHBHHHdH+%(tHUHAWAVIAUATSH8H}L&.HU1LdH%(HE1HUHtL`L赼HHH=TmIHHIHHUHH螾uIcVH}L HMH}LLaHEHLHMHHL}H}I9t/Mtbx ȉuOIcVLLL}u8AxAuL L8xȉuHH]H}1HEHEdH+%(tYH8H[A\A]A^A_]H=| 'UHAWAVAUATSH8H5 dL$%(LeIH HuH5' HoIHu6H5s HIPI9t1E1E1E1HU1HU-H5 L+HEHI9H5 HIHtH5 LF+IHu1E1E1E1H}I9u)HEH5b LHHEHEHEHu(yH5= HQÅu1E1E1HurHUH5  LI$H5 臿Åu L蹼+AAM95H1HUH5O L'LuMtH5? Lun$HUH5! LLmMt>H5 LLx_I$H5 ξÅu Lx9Mt2萻Hu(E1LZ1E1E1E1HMHME1E1^HuH IT$H5OH81̺1E1E1E1HE1LH}H} LLHEdH+%(tFH8[A\A]A^A_]UH=m H= H`y aHq H5z H=! 6xH! H= HM HSH! H=8 xH5( H=! tH= zH= QfH=y H= KH= "7H H= ~ H HsfHnHH= flfHnH{) fHn~ HUfHnfl) ~ fl) ~ fl) H5  H= wH=  cH= :OH H5 ~N HH\ HmfHnHTH= H H flH= H  >H5 H=n H=Z H=F ]UHAWAVAUATE1SH(Hs dH%(HE1LLLLLHt)H9PEHd H5MNH80EH=wH= 蛹Hd H4(w‰H=dHB H (H=dfչH. H'H H= H5If諻'Hd H0HHu1NH Huk'HHSHHAL :fRHMPHeP1舵H 1Hѹ2'1H=hyHb H'1H=kh˽HL H&&IHu i H=Ne脽HHtLHIċxȉuHڻMtL轳A$xA$uL賻t1N u轵H6&譵Ht%&H5$ 1 H &HH5 1H H%1H5 HH> H%HH%H5dH%Ht% HR H5mdHu%f%Y%H Ho =wHU =wH; =wH! =wH =wH =w$$H=`'H5^ H&HIH')'H= E1L- L,/HIH&H5 HHHH&AE&L\ H m H1H5I שHIH}&&L L H= H  H" 5HHH3&LA$={&H5 H= E1HL%m&H= .HHHe&H5< HHIHC&P&H=X E1L%n L-HIH &H5 HXHHH%A$ &L2 H H1H5 eHIH%%L Lj H= H  HP 4HHHZ%LAE=%H5+ H= E1HL^%%L ! L H=V H  H( 3HHHk%H5 H= HM%Z%L Lz H= H ' H #3HHH1%H H=wH5 H=' Hs$%L F L H=+ H  H 2HHH$H H=wH5 H= H$$H= 6+HHH$H5 HHIH$$H=x E1L- L*HIHL$H5 HxHHH*$AEL$H Hڿ1H5 茥HIH#+$L L H= H > H :1HHH#LA$=#H5j H=C E1HLf##L h L H= H  H 0HHH#H H=wH5 H= H|##H= H)HHH#H5 HHIH_#l#1L% HH=z (HIH&#H5 HHHH#A$&#H5 Hڿ1H裣HIH" #L  L H=| H U H Q/HHH"LA$="1H=_ HHH5^ E""L L! H= H H .HHH"H H=wH5 H= H["h"L  L H=2 H K H4 G.HHH?"H H=wH5 H=K H""L L H= H H -HHH!H H=wH5 H= H!!L  L H= H E H A-HHH!H H=wH5 H=E H\!i!L L H=I H H ,HHH@!H H=wH5 H= H!!L ) L H= H ? H ;,HHH H H=wH56 H=? H  L L H= H H +HHH H H=wH5 H= H] j L 3 L H=` H 9 H2 5+HHHA H H=wH5  H=9 H  L L H= H H *HHHH H=wH55 H= HL = L H= H 3 HL /*HHHH H=wH5J H=3 H^kL L H=w H H )HHHBH H=wH5 H= H L G L H= H - Hv ))HHHH H=wH5 H=- HyL L H=1 H H (HHHH H=wH5 H= H_lL Q Lz H= H ' H #(HHHCH H=wH5V H=' Hs L L H= H H 'HHHH H=wH5 H= HL [ Lt H=H H ! Hz 'HHHH H=wH5 H=! Hm`mL L H= H H &HHHDHp H=wH5 H= H L e Ln H= H  H &HHHH H=wH5B H= HgL L H=_ H H! %HHHHj H=wH5 H= HanL o Lh H= H  H %HHHEH H=wH5 H= Ha L L H= H H $HHHHl H=wH5q H= HL y Lb H=v H  H8 $HHHH H=wH5f H= H[boL L H= H H #HHHFH^ H=wH5 H= H L L\ H=0 H H2 #HHHH H=wH5 H= HUL  L H= H H "HHHHX H=wH5 H= Hcp蟝HIHoH H5 H6QL W L H= H Hv !HHHH H=wLA$=H5 H= 1HHL L~ H= H + H '!HHHH H=wH5j H=+ HwL B L H=O H HA HHHlH H=wH5 H= H3@L Lx H= H % H ! HHHH H=wH5 H=% Hq>HHHH H5 H՚H H5 H跚L L H= H N Hg JHIHnH=q1H5 H=S HL3A$`L y L H= H H0 HHH9H H=wH5N H= H L L H=k H D HM @HHHHF H=wH5 H=D HL L 1H= H H] HHHH H=wH5# H= H贘YfH= VHHH^H4 H5 H"@M-HHHGL%9AL%9E1tIHHHIxHHHHHHNL- AE=wAEHw E1LLHLLHIAEE1M"H50 LIHAH5 LIHAEH= L5 H&HE1LHE1LE1L軍HHHxxHHH5 H=d IH0H5| H=m HAH5 H= IHH5 H=% HiAH.H1HH1H1HH 1H=wH 1LHHH HHHINjm1HMrL%33AHgH[1TH= tHH= tEtLH=;AH=k Ht<1H] x-ȉu' HuH1 H5O;H8H=" 8AE$L蟑A>L臑1AELLn?ȉMZCLm~Le~H2eH=:@HHLH=wH= H x ȉuH}H1H}1HH}IxHHHALxANL`AHo}Lg}H[}111HH{xHHH/ȉH|H= L= x ȉuԏH =wH 1LHHH HHHINjxȉuHoMuL%04AH= L= x ȉu4H% =wH E1LHt HLHH<HINjxȉuHΎMuL%105A_H=[ L=T x ȉu蓎H =wHp E1LHC HLHHHINjxȉuH-MuL%/8AH= L= x ȉuH =wH H LE1HLHHHIċxȉuH茍ML%.9Aȉ]ȉHFȉH/HBzE1111LLI|$xHHHL5 A=wAH= 1HHfIn )~LHHyHxȉuHoL50 A=wAH= 1HHfIn# )LHH5yHt!aȉWHJ1yHx111LNI|$xHHHDL%-A@ȉuH聋hL%7/AȉHSL% /AȉH%L%.AA$LA$L܊ȉHŊL%{. AVȉH藊L%M.A(A$;Lg.A$nLNaȉjH7]L%-AA$LA$LȉH׉L%-AhA$L觉A$L莉ȉHw L%--AȉMHI@L%,AȉHzL%,AȉHL%,A~ȉH迈L%u,APȉ5H葈(L%G,-A"ȉRHcEA$LJȉH3y AULȉHL%+lAȉ[HʇNL%+A[ȉH蜇L%R+ A-ȉHnAELUȉ/y A$vnLaȉH|L%*CAȉHՆA$L輆ȉH襆y AUOGL聆:ȉbHjUL% *AȉHL%&AȉHL%& Aȉ6H)L%&VAsȉH贂~L%j&AEȉH膂L%<& Aȉ5HX(L%&s AȉH*}L%% AȉHL%%= Aȉ4H΁'L%% A_ȉH蠁|L%V% A1ȉHrL%(%7 Aȉ3HD&L%$ AȉH{L%$8 AȉHL%$ Ayȉ2H躀%L%p$ AKȉH茀zL%B$` AȉH^L%$ Aȉ1H0$L%#2AȉHyL%#Ay A$Lȉ HL%f#vAAȉaHTL%8#jAȉHTL% #AAȉ H&L%"AyzH~mA$L~L%"AeȉH~L%\"5A7ȉ>Hx~1L%."}A ȉHJ~L%"AȉH~H H51 HHR H5 HÀ}H H5 H襀_H H57 H臀AH H5 Hi#Hr H5[ HKH H5 H-H H5 HH H5 HH H5c HH H5u HoH H5 HQH H5I Hy3H H5 H[Ht H5 H=H H5 HH( H5 HH H5C H~H H5e H~H H5w H~aH H5Y H~CH H5[ Hk~%Ht H5] HM~H H5O H/~H H5 H~H: H5# H}H, H5 H}Hv H5 H}qH H5y H}SH H5+ H{}5Ht H5 H]}H H5 H?}H H5! H!}Hr H5 H}H$ H5 H|Hv H5 H|H H5 H|cH H5 H|EH H5 Hm|'H H5 HO| H H5 H1|H H5 H|H H55 H{H H5 H{H H5) H{sH5 H= H{UȉHxL%M,A(HUdH+%(tvHe[A\A]A^A_]UHAWMAVIAUIATMSHH=@ uHUsH2uHU1LhH@HÉ1fHnHp(HxpMtA=wAWLs K@=w1L{`HK8CPA=wAMtA$=wA$1W1LchHCxHAE%tVtXH#tW=tGH=t@H H5H8axx2ȉu,H w"H 1HHS0Hy1HH[A\A]A^A_]U1HAWIAVIAUIH5^#ATSHdL$%(LeIHEtHHHHULlH}Hu+LtLH5HH H81oLwH}u6rLIsMMLHHۙ H5H81o=LxIHt-xȉuHuH}x ȉuu1HbH}bHUdH+%(ttH[A\A]A^A_]U1HAWIAVIAUIH5"ATSHdL$%(LeIHEjsHHHHULkH}Hu+LrLH5HHj H81nLvH}u66qLIrMMLHH H5H81gn=LMwIHt-xȉuHtH}x ȉu|t1HaH}aHUdH+%(trH[A\A]A^A_]UHHHAWIAVAUATISÃ??IH8HUHHH-]AHMLEfALMGtHH?LM1L9} I Njw‰IHT HHHLrHEHuE1E1E1H}ArIHtCD61E1ƉElrIHHnIHtyM1LMLLI E1HHADI MAH 5 AUATuuuPPuPPAW^kH`IL_L_xȉtL HrHeH[A\A]A^A_]f.H=ٸ HҸ H9tHf Ht H= H5 H)HH?HHHtHM HtfD=e u+UH=B Ht H=b ld= ]wf.f.f.fHG8H@Wd1~HGpHff.H HxxH|~HHDPHHHHHc1!fff.ff.HH9t'H|~HHLPHHHHH9D@C1H9D@C1띸CfDUHAVAUATSHL-ɕ IHAAELn=wAEAID$HID$ HCC$1HC0HC8HC@ID$HC HC1AtID$(HC(A$=wA$H{x LcM9ty[1A\A]A^]fDI|$PH5$ rAID$HID$ HCA4AD$0C$ID$8HC0ID$@*A$xA$uLwoHCaf.[oDfDI|$PH5 fSfH9 H5H8Jp[A\A]A^]@H=HTH{Ht͋xt:HCH5Y H= 1@fnfUHHHHHHw H:HPw:@t y`8@H@0@1Hx81Hx@1tHyhHx(Hq@HH0qdp$HqXHpHqPHpq`p 1w 1Hx7x7t?HHH9t1fxuHHEmHEH@1HEHMcmHEHMHl HHH9ufHypHx0HyxHx8H@H9 H5H8JnɸH5 H=³ 1HEH= HCHEHxHtxt H@HElHEHyxvUHAUATSHHL%x dL,%(LmIA$=A$L%F kHHmAElffInflL-u AE=wAEHuLH]HLm)E1x AUx AUA$x A$HtlHUdH+%(HH[A\A]]L- AE=`bDjHHA$x A$}H H=R1ufDHHEkHE+LHEjHE(LHEjHE%MuDLjv&ifDHGw HG@UHHHGHUH}HtQH}HUHHOQ0HHtWxtHK wfjfHs H=U1@fHGw HG@Hpw Hpff.UHATISHHGH5+ HHHHtdHSHBpHt|H@HtsLHЋHt4xt H[A\]HHEiHEH[A\]xtUH] H=jH1[A\]fHBhLHHt1Hxt*EubHEHhfDKfDUHH;HuvH5G H=( 1H H=61DH HE1L "H H5H8R1H aX1ZDHyxtHH=k ff.UHH;HuvH5 H=x 1H H=a1DHI HE1L rH H5H8R1HaX1ZDHyxtHH=ff.UHH;HuvH5 H=ȭ 1)HaH=.1DH HE1L H :H5eH8R1H.S`X1ZDHyxtHH= 0ff.UHH0dH%(HE1Hv HEfHn)EHLIHMHSHH҉ HH8RH5L 1AH cHv_^_H}Htx =H%H=*uHEdH+%(1fDHmH=wHUH5B H=# 1H=H H}Htxu!ew@H=wHHUHMHUHAHP AXAYefDHHMHUE1LJARZYH},HV HH8j~{dbUHH0dH%(HE1Hv HEfHn)EHLIHMHSHH҇ HH8RH5L 1AH cHv]^_H}Htx =H%H=ZuHEdH+%(1fDHmH=wHUH5B H=# 1H=H H}Htxu!cw@H=wHHUHMHUHAHP AXAYefDHHMHUE1LJARZYH},HV HH8j~{b`UHH0dH%(HE1Hv HEfHn)EHLIHMHSHH҅ HH8RH5L 1AH cHv[^_H}Htx =H%H=uHEdH+%(1fDHmH=wHUH5B H=# 1H=-H H}Htxu!aw@H=wHHUHMHUHAHP AXAYefDHHMHUE1LJARZYH},HV HH8j~{`^UH fHnHSHHXdH<%(H}HHEHV)EHHuHUH} \HUHuHHH}H"H HH8RH5L 1AH DHjY^_H}Htx ~H.H=7V`HuHF wHEȋwH{xt;H}HCHt xtA1HUdH+%(H]fDHE^HEH}HCHu@^fHN wHHMIHMH5hAHUV1AXAYHE$HL.I1ARHMHUE1ZYHEHH HH8jFf^x\UHAWAVAUATSH(dH%(HE1HIH7-^HHIID$w ID$HC HEH5 LHHE[RL5 A=wAMl$M9ML= H=, IWL\HH,=wHUx]HUHIID$w ID$IG H w Hn IG(A=wAMw0HU]HUHHP Lx(wHX0x wAUx AUJHUdH+%(&He[A\A]A^A_]LmML54 M9i\HHAE=wAELj HHHUUHUINjMx xtzLHfAxAuLHU[HUx;X HH=u( x _1H fH([yH[\LHE[HEHHEZHEtH~ HE1L H BH5mH8R1H6[TXZ1HfHyxHH=41#DHH=}0 1fHH=U xuH#Z1@HZSZHUH= L XHUHSHeH. LH5H81jSEDH H= HQHHMXIH=wAYIHID$w ID$IG Hݷ w HǷ IG(=wI_0gYHtrLp Lx(v|HX1j$YHMHUH= HHMVLuM5RH)@AxAuLyXAxAtHfLHXfDxt H#XVHu} HUH5H81Q]ff.UHAWIAVAUIATSHH8H}IHIEAH~kEQMXHMIDULfE1L]HULmILILeIHHLLAWDMILEHMHujXLmZL9uuHe[A\A]A^A_]H~ H~H9tYE1H~H}ILIIHff.HULIM}YH<M9uHe[A\A]A^A_]DH9uHHeHL[A\A]A^A_]>Yff.UHAWAVAUATSH(L2L>t\M~EALELjIEHF1HELMLMLELLMHuHMI9uH([A\A]A^A_]M~E1HMLHLHIXJ 0M9uH([A\A]A^A_]fDUIHAWAVAUATSHXHFH;my dL4%(LuILo@t H;0z RA=wAE1E1E11MIFH5y I9v1I9AIFJI=wHtx Hz H9CHCHƒH)H#CHHAAdIyXH)IAPHHIHI1I?HI!L)HH9HIIMHMLMLHULUALUHUHLMHHUcNHUHHt Hw H0H9EHUkTHUAx AH@IQxIApNJH<HLMLU,MLULMH@Lm1L%kLH ףp= ףILH?LHH)HkdI)ǸH9MIHCC4 LMHcHH)HH9HBH)HLHLHMHMHMHu1} LH4H)H1HHI1oIMHD fInH}~+ HEID$fl)EHP&IA$Mx A$H=/ 1LAEx AEL%H=LAx AmLH=SE1Htx 5HEdH+%(HXL[A\A]A^A_]ÅxA$kLN^fDHr H5 H8O=KCHH fDL`N LPNH}L KLMHIH@ILMA1۾L%gHH!HHr H5eH8O.HMLuMuDL%W@KCHH H>KIHH=E1K>KIULHU$MHULMLMA$LLHG@HFH@t}HXHt6HyH~1H;t \HH9uHӾL%3HH9,HuH;5Lp HUDHU뮩tHUHUJf.UHAWAVIAUATSHXdL$%(LeIHv=_A$H H9HXHtDHqH~[1HH9tGH9T uA$Mfff.HH9tHuH;Po tfD+EHEHXxL;Mt L;=9p H[HuHEE1L- AE=AEAL-ܒ @f@ϘHcfHEH@xH8L8Htx HMHtx Htx HM1H=A$THEE8HEEHENIVHH1fDHH9t0M;l uHEpE6H|ECE1Kt I9ZIEA@HFHt?@t6MXMtcMQ1MDHI9t8I;t uvCtXLLEHUOHULEIL9NKLHH9HuH;5h uLLEHUHELhxf.MeMt L;%i MmMuHDžxE1=wIh@HHEH=( 1HuHHpHEH]HE`CLpIƋx }AHEx AzHEMIh=H611L赻HEIHHEH@xH8L Htx aHxHtx 4MtAEAE6QH]xuH}wBL}Ax AHEdH+%(HEHx[A\A]A^A_]A$=wA$IT$Hx=wL,EI!@A=wAMLAyLAZx HEHEE1HEHPpH@pHUHt4LzAL}=wAHB(HEHtwHCHpHH5M HHHpIMLM9}Ax ALuHt L;r(HEHxpHPpHtx Ax AMtAx AHEHEHEHE1E1DžpH`HJHhHEH@xH8L Htx QHxHtx 'MtAEx AEMtAx AH}LmHtx MtAEx AEyH`pHhGx H}ff.MfH8?H}L}xڃu?@HEH]H>~fDH>Lpo@5{Hu=2H}L}rDH>{>fDLh>HH=UE1HEH>mL>zL>5H=NLp=pPLp=pp=pH=5HHE1HHhH`?H}HMHUHuH5C H= 1uL}DžpLHhp=Hhp$HsDžpH`HHhLfDL5^ L9t͋H{ wHsHH 71H9s"Dff. HALH9ux uAwAA$x A$?Lx HUdH+%(H8[A\A]A^A_]Ðx HWH=A$x A$A1fDID$H5 LHHYIMIhr5IHC8HHQLx H\ H9C=wHHHU>5HUHIx x IFHHH=Y\1w1LLH4H<Ax AYAx A9HSH;8\ H;\ H6[ HJH5FH:H811x Aff.1fL7HmDL7H6HXH=A$ A$>@ff.H1}Af.HDH=GJA$yHHEd6HE8HHU5HUHHAxAff.xuH 6A1nfDL5H5~H@H=1fMAAlA`L15Ly5Hl5kH_5MLR5LE5;/IAAL5A!At/AAt8Hf.ILHU4AHUy@LHU4HUAAtI߅y].HuHW H5H85Ax AA1LLr3HHm@HW HJH5HH81-xthA1SAAL3HHU3HUuL3>H3Hv301LHU`3HU7UHHHt,HHЅtHHAX w]+HHt;x̓u2HKH=ա1]@ސUHAVAUATSH@HGH5j dL$%(LeIHLH7HHAHCH5 HHHCIŋM x tIEH5 LHHIAEMx AESLL-x 4HHH9V H9C/AffInfl…x AH=Px Hu)EHٜIHH5 H>/IAEMnx AEfInfHn~Ɓ ~ v H flflHE)M)EC @u tEAD$ @-EHSID$H}HT!Ht:x +A$xmA$udLHEn0HERA$x A$uL?0@ff.iHH=]1HUdH+%(H@[A\A]A^]fDx AEx njH/L/L)E/foEf.Lh/HHET/HEC)HH/ )I(IAEL.fDAAyL.l@LH],IŋxuHu.MtLHX.LH.dH-.,UHAVATSH8HGdH4%(HuH5~ HHIMfID$H5] LHHHA$Hx A$GHCH5 HHHIMx LL5Xt ;0HHHQ H9CA$x A$Hʅ fHn˹~H{ HEC fl)E@u tEHCH}HP6Ht1x HUdH+%(5H8[A\A^]fxtkHmH=ț+1LH,H8,L(, HHE,HEkH,fD%I %H&]A$PL+CfD%I/A$y LHU)H‹xuHHUi+HUHtH*)UHAWAVAUATSH(H]L%LP L9L-|O AE=wAE}ȿHUHuU+IHA$=wA$Mf H< w H& IF(1LMn0H=r IHqH"q MpIGA$=wA$ILJAx AOoEAoE AoE0AoE@AoEPAoE`AoEpAoAoA oA0oA@oAPoA`HtC8HCHP8IHIpx oMpHIHoCPS`oShoKxMgHIHEAW`IG@EAGPAGdA$AWhAOx=wA$LceI0MMopIILJJ!AIIGxH9sH HH9s H8xIMAGPM9I})HHIPk)IHHH HAHx A5HL H9BEHBHHHH)ȋJHII x  cIMwPM9I}(IH]x <L0@L-K AE=Y[Dx AADHE1H=Aff.x A MHtx H(L[A\A]A^A_]DLHU&HUHK H9BHHU)HUHHHUHE#H}HUIƋHUm&HU@HX&HH&AD=wI#f.H& L%L%%fDHUHUHIfDHH)HHWHHHU!HUI@L`% AEx AEHH=E1,A1AxΐAt0HpDH=MAE1fDL$fDH0H=mAzAnL$a@u9=BJHH ILR$ p<H=Γ1BJHH HIAHEMIxHEIA=MAIAHH6H=sHE1IxHEIA=vf.UHAWAVAAUIATSHH Aċ=wIc"IHtiH;QH H@t H;G LHIAMtkx A1LLAxAtWH?H=xtRHD[A\A]A^A_]@xAuL"fLx"fDLh"pHX"fDLHI2fff.UHATSH0dH%(HE1H;5'G IHH~HF wI|$x ID$HCH,HH5t LC ID$H5s LHHIMvA=wAH{H[(=wH=~ Hu1HLELEH] LEIAAx 9Ax AMA$xA$uL ff.HE =w H E HE HUdH+%(H0[A\]ÐLEr!LEHIt0HH;LEHA$x A$gHAx A3xAuLHRH=1SHEHEfDLHLELE fHAC H5RH8 v1y HtHHHEEHUȋ x t1AH H^_H}Htx HTH=1DHuH6=wHuH;5B HFt H;)B ]L[HHox HB =w H oB HcB H}Htx HUdH+%(8@H=wHHUHHMHAHULUPAXLUȃAYHufHA =IPfHHMHUE1L"LUAS迋ZLUȃYjHuHH$@ HH8jfDHE?HEHHH? HSH5H81HxH=51aUfHu fHnHHAWAVAUATSHhdL,%(LmI)E~ HEfl)EfHn)EHYLIHMIIMXIH=wHHULeHUHSJ4MLSNAZA[tsI~-@ff.ff.IIK<uH> HHL AH 'H5RH8AU1FAXAYLLeH;Htx HI9uHѻH=1HEdH+%(+HeH[A\A]A^A_]f.IH=wH~HU=wHVH}=w H}HULeHGL}HGHHv8HH)HHvHtIf.GAI)LIVLLuHHH=a H‹x LL-` fInAEfInfl=wAEH=p Hu1H)EHAEx AEmH,L;5= t9IFH; = :HLOHHgx 2=wI݉x rLMLeI}Ht xtbIM9u@DoGII DHLeHUE1HLSZY7ff4fDIu:HV=wHUHV=wHU#fH!; HH5L FAH HhH8AU1Le^_@DoGII IfH߉||H} HHHH3e w He HSHHH=$p 1uIƋM~x H5d LHH=wƒuHAxAuLLnIHtWH=c H IAEMxAEuL1LHAxLAHH=WxuHr1,HȶH=1L8+AHuH8 H5WH8FfDH`IfDAt3H;DH=*1wfDHKH1HDH=P<HH=̈́1.fHHH 8 HmH5H81 H H=E1dAEWLJAHCH=01}AxڃAuLH@`HmHH]HHOH^8 H9Cu5HIŋH8HHHufDHt H=a@ff.Ht xt fff.UHATISHHGH5Ki HHHHttID$HCLHHHtwЋHt8xt H[A\]@HHE$HEH[A\]xtEHmH=H1[A\]fLH HEHfDUHAWIHcAVAAUATSHHL%{Y dL,%(LmIA$=A$L%NY HHffInflEL55 A=wAHuLH]HLu)E>|x Ax AA$x A$HLwуHUdH+%(HH[A\A]A^A_]fDL54 A=56fDHHA$x A$HH=ҿ1sHHEHELHEHELHEHEHHEHEM`pf.fLhI ff.UHAWAVAUATLcSHHHdH%(HEHG=wI1EII(IHIcHEH~nHE1H5U H<f.ff.ff.ff.Hw HINHHH9uL;-3 EAE=wAEI}UI] =wEE1E1EH9T H^3 H9CtH\LMkI;V~=wIFHHG@HFHcd~mH1ff.ff.ff.ff.IHTIHTPHHtHHHH9uÐff.UHAWAVIAUATSHH HdH%(HE1H=K  HM H{H9t>HXH HqH] 1fHH9C H;T u=wH;)  HLAoEIFH;") ) t H;2(  A=wAHDžE1IEE1DžHDžDžHHIFH ' I9N I9IFJI=wMtAEx AEH#L-( L9{HCH5_ HHH HH L=' L9H;' @@u L9Yx u HDžHCH5o_ HHH HH L9H;Y' @@DL9;HHH  x L HDžHCH5^ HHHg HHf L9H;& @@u L9x jHDžHCH5L^ HHH IM Ax A HCH5-^ HHHIM|Ax A HCLHH5] HHwLHX0x 0iHHy@H1L9HLLM9. HHFIHLM9hHHHE1HHHHPHIHHIHH9HD\H% H9Cg HCHƒH CAI)LIQHHHq@HM]L9]HcLLհHxH(HH(IHHLHH HHHtH=# H0H9Ax AHG HCH9t5HXHq HqH 1 HH9u H;T uH|Hfo HHHx$fo0D$fo@D$ foPD$0fo`D$@fopD$PfoD$`foD$pfo$fo$fo$fo$fo$*HHHeH;x# MHPF HHHH9HXH4H~HO1 HH9>H;T uMjHHHIIջA`AEx AEHH=91pHtHx MtAx AHEdH+%(HeH[A\A]A^A_]Ð[H HIHHw@HIH^! IA2H5J H8hHKDH="nH蔬߻H-@H! H9BHBHƃH H)rHHH H7fHcHDŽ0HHDŽpHDŽŰ@LHff.ff.HH9HuH; ff.LHL~HDžfDHQ H9B HBHƃH= H)rHHH HDI9K\ I=@HHH_IIջfI94M9HDž1HIHIE1HHHcLpH0HŰIHcLŰHI@HLHLHnHLIH0H@HHHt I*IH('Lh[fDH9 H9B HBHƃHz BAI)LLHH@LH@@H~HDžIE1HDžHHH9HOH+sLhsHXsd+H=> 111cjHH=fDLLLDDHHH(fI߻AH"ALHHHI߻fDcHLH IHHIAALzDH? HH9FHXHHqH~"1ff.H;T HH9uH} HJH5gHWH81LIE1HDžHAAH:H;HDžIIH1LM9}HI߻:H%H;6HyHFIHLM9DHH)HHmHDkHCIHw@HI HALXI߻~AI-fIIջIHH9HuH;  fo 11MH$fo0D$fo@D$ foPD$0fo`D$@fopD$PfoD$`foD$pfo$fo$fo$fo$fo$HHH H;& hH< HHHH9KHXHYH~Ht1 HH9cH;T uMfDHHiHHHHHHHHH=HDkCII IHH1HHIHI9%HIH)HHHHI[HH1HHIHLLM9HH)HHHH{HH=15E1H1DI߻s6IaH^I߻MLHDžE1M9IHHHH+HHHHHtUHHHHHODHeH,HmHH[HDžI90LIHDžLgHH)HHHoBrHH HHHH}HDžIHIHH)HHH\BrHH HHdHHHDžHHHHItHHwHHAALHLHH)HH.HDzBII LmHHHH$BrHH HHI߻`H H5ʑH8RqHH~HHBrHH HHmIIHDžLH IAuH59 H8HHHHUDzBII IL8HHH9HuH;5 vffE1HKHj H=M距LHc H1MH5@_H81nH& H==v#H" IH5b< H8H8H=^H,߻HHG@HFH@twHXHt3HJH~1H;t HH9uMH<IHH9HuH;5 tsHHH9HuH;k MHBHQIMH5u]HH H81HH H5ŽI޻H8BHjwHHH9HuH; M8HBHQIMH5\ HHr H81PHH H56I޻ H8Hތf.UHAUATSHH;52 HHsdIH-H9 H9H@HM` A$=wA$Mh(AE=wAEAx AL;%D L;% u I9?HCLHHHSHHR(HHA$x A$<AEx AEHH[A\A]]fHEH H5ZH8VLEAxAuLHcH=[1豙HH[A\A]]LH @LH H=E[1^f=wHHH[A\A]]ÐHLHHfL8HH[A\A]]f.LNHHHLEH HDHl H5ZH:H1LEHA H5ZLEH81OLEfDfDfDUIHHH f0 LOL9t5IIXHt^H~H~w1fHH9taH;L uA=wAL;! uoAMxAtAL@LDHH9tHuH; tLLLfDLLULULH/ HL9xIXHHqH~1@H;T NHH9uH HJH5jXLEIQH81LEAxAtjH} H=2Y͖E1H~ H5LEH8LELHH9HuH; _L댐UHHdH%(HE1HQHHPHHHHHoHcwd) o@)0o@ )@o@0)Po@@)`o@P)po@`)Eo@p)Eo)Eo)Eo)Eo)Eo)E~pH~H0HHHPXH9H)HH9P@uHHH9t5HxH wHUdH+%(H vDH HE1L H H5PH8R1ḢXZ1DHyxkHH=|d1pDH(sH=WHHqHH@UHHdH%(HE1HIHHPHHHHHoHcOd) o@)0o@ )@o@0)Po@@)`o@P)po@`)Eo@p)Eo)Eo)Eo)Eo)Eo)E~hHHPXH H @H9PPuHPHH9tI HII9t*HGHgLGPHIMHHI9uIK 3AfDDHD)x~Lcƍ 7IHcHJNH I)fff.ff.ff.H0HHHqHpHHqXHHL9uHHcH Аff.H@H0HHPHH9uHpHH>HۀHc0fSHHH LLE1HHH ףp= ףLLHJ*LH?HH)HkdI)EAEIEIŸI9C4$ICHcHL)HH9HBI)H#HHM|LHMuA HLH4H)MH1HHI1pHHH+HME1H ףp= ףIJ*LH?HH)HkdI)EAEHIŸI9ICC4?HcHL)HH9HBI)HFHHMtLHMz1A HH4H)My F-HHH1HHI1GIMLH]! fHnfIn~ % H~! H IVfInHCflfl~% )HT.IT$fl)H)GIAMmx Ax A$x A$H= 1LDCAExAE @ff.HXrH=sAH蟀$D)ރxvHcHHcHHD= H|=HTH)fDff.ff.ff.HHHHJHHHHJXHHH9uHcHEHLf.H@HU`HHPHH9uHHPF-HHf.ILHLfDHHH9t&HBHt+LZPHIMHIH9uLN7DLM9s H9AoEHE{HDžA)oE HHX)oE0)oE@) oEP)0oE`)@oEp)Po)`o)po)Eo)Eo)Eo)ELHcJ4HMHI)I8I!f.H;P@uHHL9 HxoEHH5H)oE )oE0)oE@) oEP)0oE`)@oEp)Po)`o)po)Eo)Eo)Eo)E!fH;PPu*HPHH9Hxff.AFtZDLEHHu`ALAVH}H0L,uY^Hn1uH[HNL#.I>1ɋ7w"H7L9>II>7vHI9u LDHHH0AE1HDLHEDKHHXI9VH>I`HPLCH6HHuJ7HH9s J6H9ZHEDKL] LcHDžHHXSHJ4fff.ff.HQPHQHH9u{HHH~ICHHtLXMH5MfoHDDA$fo$fo$fo$fo$fo$ fo$0fo$@fo$Pfo$`fo $pfo0$fo@$foP$fo`D$fopD$ foD$0foD$@foD$PfoD$`foD$pfo$fo$fo$fo$fo$HĠtb=wHHUdH+%(uHe[A\A]A^A_]SHAxAu LHs\H=.j1@LqLغH HHI~H9aHXH^HqH~1H;T 3HH9uH= HJH5+HWH81 ff.;fDH9 HI|$H9HXHHqH~1H;T nHH9uH HJH5 +HWH81|ff.fD諳I蛳IE諳HQSDH H5\H8袺 H H5[H8臺HfHH9HuH;, fHfff.HH9iHuH; WT@UHAWL~AVAUMATSH(H}LMHL9nHBLIHEIH]H}LI0HMH}LIL4LHLL׺HuHLL)źL;}uH([A\A]A^A_]L9MHZMMfff.H}LI谴HMIHEHHI$HHI$M)I9uH([A\A]A^A_]fUHAVAUATSH^H9CIILjff.H1E11LεI II IHL9u[A\A]A^]f.fUfH IfHnH HSHhdL%(LUI)E~ HEfl)EfHn)EHLLIHM<IMtI H=wHUHH]ML]H?ZHUHPJ4LU%_AXtuH}LUL]*I~IIJ<uH HH YH5!L _AHYH8AR1袯Y^HH]H8Htx HH9uHYlH=<,e1HEdH+%(H]HÐIt IHHV=wHUHV=wLHUA=w AHUH}LMHH = H]MA=wAHIsH AQLjhj RHjH=> LU LUHHPAHx AWHH]H8Ht xtHH9uDHMHE[HMHE֐HEGHEnfDIH WAH HHWL ]H5H]H8AR1衭XZf.IuHV=wHUHV=wHUfH EWAmfDH =wHMHH]BLMHUH}H e 'LHM4HMH 9 =wHMf.xAt%HVH=i)HMbHM8LHMòHM8UfH( IfHnHAUATSHHdL%(LEI)E~h HEfl)EHLIHMIsISMHLUH]LeHBVJ4HLPC!Y^LMLUMgHuH H=wHUHfIIMMH 8UH? HH]L gZH5LeH8AP1IH{UXZH;Htx HL9uHTH='`1@HV=wHULA=wAHuLMH~H H]LeMAE=wAEH  H HARH= HIrjQPjQLPj5 ~ HAEHPHBx AEH;Ht xt4HL9uHEdH+%(DHeH[A\A]]f.HMHM뽐HV=wHUH=wHUL  A=fDH 5S1fL A=wALMyLHMTHMHY =wHUHH]LehfDx AEt%HRH=%HM^HMLHMʮHM?@ff.UH fHnHAUIATISHHdH%(H]HHEHV)EHHHUHHUHHH(H HH5L :ZAH OH8R1HO谧^_H}Htx HQH=$]HEdH+%(He[A\A]]H^I] =wH]=wI|$x  I\$HCHH5 HH)IMH5QL輮H5PL襯HHoAD$oBAD$(HB ID$8HCH5 HHHHHI$xtZI$1AEx AEtOH}HXKfD;fL(H}HuSf.IM wHHMHUHMH5MIAHV1AZA[H]<@HL-nMI1AUHMHUE1HZYH]HH< HLL WAH LH5H8j1AXAY?DHNH="ZfD IH =wH HuHHHEHEIċxuH說Mt%11LA$xA$uL耪H;NH=M!8ZfDHEoHUHIHMH=!YHMH= YfDHH8aUHAUATSH8dH%(HEHGH; =wIfHnfH=@ HuHfl)EIHH5 H*HHAEx AEH]fIn )EAD$ @u tEC @EIT$HCH}HT.IA$Mx A$taxtDHEdH+%(jH8L[A\A]]LPvH8fDL(fDH; H}PXH}IM\ff.HKH=E1WLA$x؃A$uL跧DA$x A$tWAExAEuL膧@x A$t5vkHS^fDL@L6@H;a H5$ H}H}I肥fUHAUATSHHHGH5t dL,%(LmIHLH1HHsHCH50 HHHIċM,x fID$H; A$wLA$x A$]IIEH5' LHHHHHCH5 HHHIMrx IEH5N LHHHHAEx AEHCH; wI݉x LAD$ fHnfIn  )E)M@u tEC @EIT$HCH}HTDA$Hox A$x HUdH+%(HH[A\A]]A$x A$-xuHFfDHGH=mS1@8HH"LSLأHȣeHHE责HELHE蜣HE苝H{IH;Y LLPXHA$HA$L(A$A$DHۜI˜HA$xA$uL觢AETAEGL:f.H;Y utHPXIMfDxA$uL3IL&H;H H5 L˟HH;$ tH5 H諟IuNff.UHH#Hu^H wfDHQ HE1L zJH BH5 H8R1HKE XZ1DHyxtHH=(E 1@UH fHnHH@dL%(LUIHE)EHLIHMHHxH HH8RH5j L L1AH "BHDH^_H}Htx $H7DH=4P1OHuH=wHUHB u@HI w x  iHUdH+%(@IBHULH5 HHHUHHDHGH5 H}HHH}ȋx u腟H HU؋wHQ;DH=wHHUHHMH(CAHULUP\AXLUȃAYuHUfHHMHUE1LBLUASZLUȃY:HUH~H HH8jfDHHE蜞HEyDff.H;BH=8NHU1DSfDu̓^_H}Htx 8H=H=E1IHEdH+%(7 HeL[A\A]A^A_]HbH>=wH} ]AŃ1蹛IHHEHX=wIcϘIHfHnH= HE1fHnHHHxfl)EIƋx wAEx AEMiIFH; t H;e A=HEL1Ax AVIE1H}IFH2 I9VHH9JIFL$HA$=wA$MtAx AAHMHy=w HyH H9GHuHEHLeH};LUIAx AMIGI;G AuwAuIwL,HIGAEx AEzMH}LULULUHI LUqLUHHt H H0H9}LUyLUAx AMAx}AuuLkAxAu LٖMtAEx AEMtAxAu L裖H[:+H=]E1UFH}H~^tfLHMwMFA$:A$ASDLyLH=wHHUHMHUH9HAPAZA[H}f.H舕|Ax AHMHE1H9Md HA$=f.HHMHUE1H8S$ZYH}H+H HHL @AH (6H5SH8j1GAXAYtE1MLxA=AMAEL_fInLWAfInfl=wAA=wAx HxL׺L]LU)E)L]LUIAAL诓LUfDLLEMA$A$ALPYLHuIH=wAIGH5P LHH HH'Ax A4HL4INjMCx L;= L;= #L;=\ L&ÅAx AMA=wAIvH`MLH=  IHAAL豊@H萂ADžU#Ae0L% A$=wA$LeH =wHUDL%Q A$=wA$LefDHI =wHUtf.L% A$=wA$Le;fDÉ'fDH谉IHV=wHUHV=wHUDH`SZfDL@L0HH ,H,dH=E18H= HxLHxHH H LH5P*H81,Aet+ff.H ,DH=7E18bfH(fDIaAeAxAuL臄H` HH5'H81/Ahff.UfH fHnHAWAVAUATSHH8dL$%(LeI)`H`H-`)pfHn~qv HEfl)E~Qv fl)EfHn)EHt+LIHM~IH3JcH>@IH3JcH>DHV =wHUHV=wHxHV=wHpHV=wHhL&A$=wA$L`L- HhHpHxH@HyHEH`H0HHpLhHA$=wA$A=wAM9L= A=w AL= H@LHHEHEHEHHHHAx AcHH H H=7 HQHH 蔂IHX=wAIGH5 LHHIMAx A HHL޺L L HIAx AH  L;= I9H u M9Ax Ax`H H=2 HQHH菁IH=wAICLLH5 HHLIMAx AHHLs~IHGAx AL; L; u M9 DAx AE H0H;5ե H;  L9 H0yA7H8LA=wAH1LLASH8L0HLH AZA[HL0IcA fHV =wHUHV=wHxHV=wHpHV=wHhH=wH`HHUJ4ML5>$H`AVHPA_ZHhfHpHxrH}M~DIIJ<uHl HLL +AH 5#H52H8AT1&yA\A]LeH;Ht xtIHL9uH #CH=eE1/HEdH+%(BHeL[A\A]A^A_] fH0H;5r AH; DL9H0vA5 H8LA=wAH1LLASH8L0HLH YL0HI^ Ax AW L9:A=$ HHMAxd ff.Ax AMMtA$x A$Ax ALe$ff.ff.HL9^H;Htxuy}LXuAAx AgH!DH=^E1,HHx FM#E1DL|JL|oA$=wA$AEx AE L= A=w AL= A$x A$^ MMMbH  AL M%H HHY H5H`H8AT1u_AXf.L|H5Y H =wH H`HEH0fDH  H@=w L`H@HxQf.AE=wAELpfAE=wAELhfE}HxL`L-3 H@HEH0@H@H;j t H;ٟ H  H= HQHH8zIH =wAICL8LH5 HH L8HHAx AH H9A HHHHEL}H8mL8IAx AGMICL8LH5B HHY L8HHAx A<H5 HϺH8?vH8HIË x L; L; u M9GDAx AtEH@=wH@H5ϝ H9pR L@HL}HHEL!IAEx AEq AMt^HHHHHL@|xL@pXA?H DH=g (HHLL xL DLw6LLHwLHDLwH  H =wH HE@L- AE=wAELhuH =wHxmL-Q AE=wAELp3HwADH AL #LvHvE1AbfH H= HQHHuIHA =wAIGH59 LHHIM4Ax AHHL޺LrLHIH Ax AL;=m L; UM9LLmAx AQH0H;5 H; 0L9'H0\mA H8LA=wAH1LLASH8L0HLH L0IXZMMAfLt+uH H= HXrLXH MH@AnH?H H@H5DH81 nLLIlLAAE1ff.MAALtLL8kL8AAHL8sL8TLsfDHHL@sAL@0mD0HAE1 WmICmI5ADL srLsHH AMLrLLrLLL8rL8LH8rH8)LsrLH8_rH8H t H=% HQHHqIHf =wAICLLH5 HH LIM} Ax AHHLfnIH Ax AL; L; rM9iLLSiLAD Ax AeE5H0H; AH; D\L9SH0hAǃj H8LA=wAHE1LASH8LL0HLH AXL0HIAY0AOMApHH=? HX{nLXHMH@HjIH E1AfDAiLI4pH8H= HXmLXH8M H@iHH H@H5SH81/ijf.A `iL8HDLIfInLAAfInfl=wAA=wAx =LHu)EL0L8.L0L8IAALL0nL0L8LLnLDthL8HL]ndHPnMHLL@9nL@tLnhADH5! H=Ҵ HVH2mIH=wAIGH5 LHHIMAx AHHL޺LjLHIAx AlL;= L; _M9VL eDAx A<kH0H;5i AH; DL9H0dA~H8LA=wAH1LLAWH8HLHe AZA[HHAx AIlHH= HXSjLXHMH@AfH]@0eD0HAE1oAEHH@fInLifHnH8HƋfl=wAE=wAEH@x HuL)EH8IËYNHL@ kL@3Lj*LLjLBLjHL0L8)jfoL0L8L= H=^ IWLiIH=wAICLLH5r HHxLIMBAx AdHHLfIH; Ax AL;P L; -M9$LLaLA Ax AE H0H; AH; DL9H0+aAǃ H8LA=wAQE1LASH8LL0HLH ^L0HI_uA@ iHH= HXfLXHMsH@bIH E1A@A%0jbD0HAE1'bLI2AAE1EELLgLSLgAiLgH) gfo >DaHAE1H5 H=; HVHfIH:=wAIGH5M LHH IM Ax A HHL޺LcLHIAx Ai L8)…L Ax A H0)AH8LA=wAW1LLAWH8HLH AXAYHHA `I^MfHH=׬ HXdLXM _HAj_LILeEaA+AJeH=V HXLcLXMj_HE1A0H_D0HgAE1Hc H@H5AH81^[cALLdLLd*H=ɼ <HH5 HHLHIxAx AHHL`IHAx ALL&LAAx AEH0k&AǃH8LA=wAPE1LASH8LL0HLHd L0IXZMA4LH8UcH8H LH5H81\H HH5H81\@PcHH=ک HXaLXM\HteAqA0\D0H0AE1=Lb)LrbH HH5H81[w^\HfAE1H=H IHH5 H`IH)Ax AHHL޺Lv^LHIAx AL($…Ax AH0#AtZH8LA=wAP1LLAWH8HLH Y^HHA0[D0HtAE1H5! HH-H5HH@E#L@A/Ax AEL=H A=wAH=a L9HHH@t H; HHWHHHE1LHMHH@LMLH@Iŋxt]AxAtcMt11LAEx AEtAE1VL_HH@VHXH_ALk_L-2 AE=wAEH E1LHHLUHEuIAEx AEAM11L"AxAtAE1L^AfL^AxL^E1A1Ly^AE16L^^LALL?^LRAE1L^(ALL]LAE1AH2 H@H5AH81dW_AJAE1fWHYAE1*uWI`LL^]LaAL?]f.UfHد fHnHAWAVAUATSHH`dL%(LEI)EHhH)EfHn~M HhflHE)EfHnfl)EHt.LIHM~"I H= JcH>IH/ JcH>DHV=wHUHV=wHUHV=wHUH=wHEHUHL}MLmHELeHhMHEH HUHP=w HТ HEHUHHHHEHHIƋx hMOH5x H=) HVHXZHHM=wHCH5 HHHHHSx BLHX~WHXHHx :H;# H;~ u H; x !yH5 L/H`Hw=H`MHHPH=N~ H2 IHK |qHYd@HV=wHUHV=wHUHV=wHUH=wHUHHMHHhLeJ4PLY^)H} H}L}MLmMLL-(~ AE=wAELm*DH5 H=Ÿ HVHX"XHH=wHCH55 HHHHHx LHXUHXHHx H;| H;| H;8} HP[x ?H5 LmH`HO=wH`MHHPH=X} H IHA<M7H 1H{ HLeL GH5H8AP1IHVPXHEZHhHhH;Htx HI9uH H=:E1HEdH+%(HeL[A\A]A^A_]DHNA5x qDHAE1H=@Ax AHhH;Htx 5HI9uUH9{ =wHUHH =wL}HUML= A=wAL}f=wH`MHHPH={ H IHA7L= A=wAL},H =wHUfL-Az AE=wAELmfDUfDHTTfDL-y AE=wAEHELmLeHhHHXTHXDLxTHhTNH߉XBTXH HTH4H=ME1+THXH= HxQHxHXHH`MH)A5f.{MHw=wqH`MHHPH={w Hg IH=HA>A5|HRoH`MHHPH= w H IHA>HLHA: @HHXRHX DHxR0H߉XeRXaH`MHHPH=v He IH3A9]A6^RHXH= Hx$PHxHXH%H`KHA:4A:L-s AE=wAEH= H;=v  IFt H;u L3HIMt{HHLL}HHE膼AHÅx AAUx AUtrHt11H.xt^A@]fDAEx AEtAA;@LPA;"HA9LPHPLP\L$GI Hu H`H5!H81INHu H`H5H81IW@ff.UHԥ fHnHAWAVAUATSHHdL,%(LmIHE)EHLIHMHHWHhs HH5EL AH H8R1HRI^_H}Htx HH= 1@HH>=wH}HG`HGHHv0HH)HHfH7HIfGAI)LILI\$HHIHHIEH5s LHHIMb}QIH1HiOHHH5 HL Qx H 3 H= HQHHMDMHH(=wHCH5 HHHlIMNx H5N LLkPAx AIEL=ݕ HHH=ƻFLLLHJH AEx AEAx A=wH=` 1HufHnH )ELIƋxx @MA=wAH=h 1HuLuHHE?LHAxAx A)H/HCLhpMI}L,MIHH=q HHIIA$x A$MLHAUAx ApHxfDH}Htx &HUdH+%(He[A\A]A^A_]@H=wHHUHMHUHHAPXAZA[H}fDgGII GDHHMHUE1HS ZYTH}HHun HHL AH H5;H8j1/DAXAY fDDgGII If.HEGJHEfDHPHm H5H81CxAuHJAff.HDH=1CLLLIHHfDAE11AEx AE`Ax AHtx MnAcAWL?IJf.H(IAExAEuLIADHHFHHlLHFLHCHSfD IBHkBIHXHLHH3L8H6L(HAE1mfLHHHEGHE[HG*fA1 HHuH= HUEH]HAIHE1ALpGAADKAILHE4GHE{L GE3AHHWHj H5E1AH8(H>HDHHHcH>DHV(=wHUHV =wHUHV=wHUHV=wHUHV=wHxH6=wHpHxH}&H}@LMLpLmMKH}HxHMLEHH;=Xi H;=.i u H;=i *HLSY^fDIM9t7I<$HtxuIHhqDHhM9uDHUdH+%(He[A\A]A^]HV(=wHUHV =wHUHV=wHUHV=wHUHV=wHxH=wHpHLmH4ILLpLARLͲAYAZHxH}H}H}H~WHHBI<uHf HAL .H H@H8SH51<_AXfBIM9tI<$HtxtIM9uHnH={n17LHHPHXL`HhX:HhL`HXHPLHUf.UfHh fHnH0HAWAVAUATSHHhdL%(LEI)E~W1 HEfl)EfHn)EHLIHMIc = MtIH=wHUHHH]PJ4HHU艮Y^H}S LuM HEH`HR H L} L(hE111AHLAIH; =wƒ H8 L(hE111AHLAIH =wƒx` AEA;D$L5r H=# IVL=IH=wAI@LXLH5 HHVLXHAHRxAuLHX=HXH~b H9AHuHLmHHELeHXϨLXIAxAuL^=MFA=wAHƃ L0IH =wƒ H5. H=߃ HVHX?H׉X1XHPUH=MLIfInLyAfInfl=wAA=wAx 6 HuL)ELPeLPHAALHP0HPLHh0Hhx  DžhGE1E1LHh0AUHhp HHhb0A$Hh?Gf.L80&H(0HLX0LX{DA$@AULL@HX/L@HX8HL@LX/L@LX)H:H-DH=g*E1E1/H=$v HxL]-LxM8)H HPH=D )HHEH=קk/H=u HxL,LxM(H HGGH=DLHX`.HXTHH@LX>.H@LXLHh.AUHhE1E1LHh-A$HhLHh-HhGHHh-Hh'LXHE1E1x A4DžhPxc'L@HAE1E1x ADžhGHh'&HHVH=O%HXI)L5 A=wAH HuLHHEHEݖIAx AMt11L薝Ax AHHH=H L@HXs+HXH=q Hx9)HxHXL@HHXL`Hh$L`HXHAADžhVLH`*H`HDk$L@HXIWDžhLtBAxAtCMtAxAtE1E1@L*fDH)fDL)MYfInMQAfInfl=wAA=wAAx A1LHu)ELPL@LX軔L@LXHLPAALL@HX&)LPL@HXzDžhJAMPfHnMHAfInfl=wAA=wAAx AyLHu)ELPH@LX轓LPLXHH@A&ALHP/(HPH@LXAH=> =wHv HuHHhHEHEHhHx Ht+H11HhșHhx H;WH=u8x DžhXSDžhPDžhGHLX)@'fo@LXHLP)0&fo0LPx 1DžhKE1E1HE1E1&DžhGL& ACZLi&LLPL@LX)0@&fo0LPL@LXLLPH@LX)0%fo0LPH@LXAH%DžhVHh%HhH%JH%HE1E1%DžhKHJ LH5+H81HJ LH5 H81#HJ HhH5LhH81LhAA1DH@J HhH5H81x>#ff.UHAWAVAUATSH8HHHHHpLxLdH%(HE1HDžHDžHDžHDžHDžHDž=wH=wHx=wH=wH{ H=j HSH#IHj=wAI@LLL`H5s HH_L`HH:Ax A HDžHG H9C-HHfIH)MHIAEx AE M.Hx  ID$H5Iz LHH2ILM1H5 ffo)PL9H=LG IAH0H9 AYAx AHDžID$H5o} LHH ;ILM;IRHBpH.2H@H!2L`LHL`HHH;AxAuL!H5 HDžHDžH9? HCH;0>DkAE HxH;GF 1HH H<L-1y H=g HDžIUL; IHM=wAIALLL`H5|| HHPL`IM0OAx A|5L=x H=Ng HDžIWLHHO=wHAH`HH5s HHJPH`IMOx 8L=x H=f IWL(HHBP=wHAHPHH59s HH>THPH`H`cSx 0H+D I9F UH`HLHHDžHwMHIH`x F0Ax AN0MFUIGH5q LHHWHPHPVAx A$0HDžHTC I9E`HPHLMHHDžH蝉H`HHPx /Ax A/H`e_HxH@e HDžH@H9 HXH HqH1@ff.HH9H;T u Ax Ab7HDž]H0LpxfIHt H;iB MvMuHDž`E1L=5v H=d IWLjHH)=wHCHHH5q HH7ILM*x ,A$=wA$H1H=q HHLHDžHHHHA$x A$1H)HA I9GL7HLHHH HDžMH MHËx ,Ax A,HDžHDžH6MtAEx AE3MtAx A3H`Htx f3H5y HHIH-H? H5? I9HI9HL;@ LL`L`A)-Ax A/HDžEuL-r H=a IULIHI]=wAIALLL0H5at HH^L0H`H`]AxAuLH`H > HDžH9H`L`foPHHL)HIAExAEuLMaH5x L95-IGH;s> 7+MoIAAx APHDžEL-` AE=wAEHh HLHHDžHPH`HAEx AE)H`L`11LAx A)HDžE1ƅHDžDž`HDž01E1E1HDžHE1HDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPHDž`DLLlHH;; +LϺL`/L`HI&H;; H;; &2L;-a< 2LL`$L`AEx AEtƅE1E1E1HDž11E1DžXE1E1HDž0HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPHDž`G"ff.CL5An H=\ IVLVIHJ=wAEIEH5o LHHMH`H`HHMAEx AE'H`HQ: H9P!VL`foPHHL)蠀HIAEx AE&HDžMTH5s L9EIGH;0ZMoIAAxAuLHDžEL-[ AE=wAEHFd HLHHDžHH`HAEx AEa}H`LL`11LgAx A:}HDžE1ƅHDžDževf.ff.HH9tHuH;7 fL-k H=Z IULIHs=wAIFH5h LHHtILMsAx Al_HxLPH5e HGHH6yLPIM[yH7 I9BZ~HoZ HLHHDžLHL0~L0HPHAx A_AEx AE_HDžHP~H6 HPH6 H9HH9HFH;5:7 FH AŅyHPx cHDžEL-i H=X IULIH=wAIGH5em LHHKIMֳAx AL5i H=;X IVLH0HH#=wH0H5d HGHHIMH0x HxLPH5c HGHHLPIMPHi5 I9C1HLHHLLP{IAx A'HPx MZIFH55c LHHIM_Ax AH4 I9E1HLMHHL{HPHAx AAx AHPH`=wH`HP HIH&H;AH;DiL;%4 \LL0L0AXAx A`HDžE!HP=wH`LPxH`HPx AE=wƒgH`x qL`@ff.HT L(hE111HxAHAHIH[U=wƒ|iuL HxHDžxuH L;=2 P!IWHMwLH5e HwXIMWH5k L9KFICH;+2 [ISAuHHSx AOL-S AE=wAEH3f HLHHDžHxHPAEx AELHPTLP11L~ALx AiLxE1E1ƅHDžDžuHDž01E1E1HDžHE1HDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžP=wLkAE=wAEHIH`1DD~LH  H=R HHU LMz0HwHDžE11LHDžE1E1E1HDž01HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPHDž`ƅDžWf.HDžE11LHDž0E1E1E1HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPHDž`ƅDžWDMtAx AMtAEx AEMtAx A6MtAx ANH0Htx QHtx [HpH=E1ouHF8<Hx Htx H`Htx MtAx AyHPHtx lH Htx _HpHtx RHHtx EHHtx 8H@Htx +HHtx HHtx HHtx HHtx HHtx HHtx HHHtx A$x A$HHtx >Hxx 6HEdH+%(lHeL[A\A]A^A_]ÐLLHLHLLHLHLLHLHLHHDHHqHDHXHHH8+fDHHHH fDH#L(HrLzfDHxHhHXKfDH8H(H L`H@Hq( HDž`HH8( HHH;( UHH;ɉ=w HHLHDžH91H=wL5t[ H=%J HDžIVL~IHO=wAEIEH5] LHHSIM&RAEx AEBL5Z H=I LPIVLLPHIS=wAEIELPLH5xW HH{ULPIMTAEx AEEH& E1AI9C]HL0LLPHDžHLLPL0HIMHT HP =wLLLL)H?JL0H LPLL0LPHIMtAx ANHDžAx AGJAEx AEMJAx AUJM;LHx JHHDžLLHPHpH;H;fH; % YHyDEHxH;% 1=wLxHH=jR 1HHHDžLHHpHAx A6LHp\Hp1ɺHlxHIH[HpHH0L0HH ^Ax AKHpx nKHH=wLHHP1HHH=|Y LHHpAx APHDžHp*dH H=wH HDžHHpHIH"cHY H oY HP =wHH=Y HLHH Hx YHDžAEx AE_LHjL-V H=D IUL'IHm=wAIBLLLHH5N HHuLHIMuAx AYL-U H=:D IULIHr=wAIBLLLHH5Q HH{LHH@H@HH\zAx A\H! E1I9FyHHHHDžLHHHH yH!O HHHA wH@LHHH)HHH?H HH4%MHHItAEx AEdH@x cHDžx !cHDžAx AcM {Hx aLƅHDžHDžHDžHDžHDžHDžHDžHDž@DEHHBB H@H9t5HXH|HqH1 HH9H;T uH=wLH1HH=O fInX )NHHAx AaHDžHH Hx AHHHID$H5Q LHHjIMiH5QX L9IH  IFH0H9MxEnAAx AZEIuuHH5Q HGHHEIMH5W L9jIFH;0EnAAx AEjA$=wA$L-)Q H=? HDžIUL3IH&=wAIALLLHH5I HHLHILMAx A"yL-P H=?? LHIULLHHII=wAIALLLH5M HL0HœL0LHHHHAx AdzH E1AI9BHLLL0HDžHL0LHHIXHI HP =wHHLLL)LHLH?LH HJ4LLH0MtAx AHHHDžx yAEx AEyHDžAx AyH0HDžHLHDžHHIHHF HP =wHHLHH=R HIA$x A$UH0x wAx AwHDžMHDžHDLsLkA=wAAE=wAEx HHLLHT`HIAALfDHDžE1E1LHDž0E11LHDžHE1E1E1HDž1HDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPHDž`ƅDžWMtAxAtZMA~ArLLLHLLH;LLLLHdLLLHae5HH\H=9 HL&HHHuH5 LH5H81qHHtx HHHDžHtx HDžHtx yH0H5 9 HDžHxpzDž[/H0H@xH8L0Htx BMtAEx AEH` H`H`LLLƅE11E1HDžE1E11HDž0HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPHDž`ƅE1E1E1HDž11E1DžXE1E1HDž0HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPHDž` IRH&HL L DLHBhHLHHxL` jL`HH=wHDžLDžH(6 HP=wH 6 HPL-~M AE=wAEL-eM Hx u$LHRHLPLPLLVLLrHe3HX,LK2H;w iLH`HH`H;AH;DgRH;  ZRL`LAAx AEJLDƅE1E11Dž_E1E1E1HDžHDž0HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPHDž`7HzOLHLHLpLPLHLpLP1ffA.AEAx A#KHDžHw3 HPH#xIWHPH9HXHHqH~1H;| cHH9uHPHRH5\E1HHH H81b1E1LxƅLHDžrBfL`iL`HHH9HuH;Q .fDLH !H DL=LwH[H=c訚H0HHHWnHHP=wHd1 =wHLzIGH; <A=wALzM;H=9 LIHY=Ax A.HHHLLpHDž=ULpIAx AX/x T/Mt$HP1L[Ax A+3Dž]IIQxHHHH;~ LXAwALx(MtAwAHLHL HHHLplHPLpHHL x &+IQxH:HHtx  2MtAx AN2MtAx AQ2HHLHt H;Z(nGIypIQpHtx *2MtAx A1HHDHL`L8HDžE11LHDž0E1E1LHDžHE11LHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPƅ HIHDžE11E1HDž0E1E11HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPHDž`ƅDžcLHDžHDžE11E1HDž0E1E1E1HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPHDž`ƅDžcU2ITAEۅAELL`L`HDžMWfHnMOAfInfl=wAA=wAAx A.HLϺHLL )OLL HHAWAKLH,HL )ƅE11E1DžiE1LE1HDžLHDž0LHDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPHDž`EH;l HH`HHgZH; AH;\ Df%H;  Y%L`LAAx AdEoƅE1HDžDždHH-IHDH;L;L;f LLp)LpA[KAx A-2ELH5t@ H1IHRH;L;Z#L; M#LLpLpA8WAx A6=EZHxH; 4L-9 H=B( HDžIULHpHHr=wHpH53 HGHHHGrIMqHpx [H5D? HxsH HH:pHm I9FpH HLMHHDžHKIH x ZHDžAEx AEZMTnH1LLpLpHIXuAx A#^L;AL;D:L;5 :LAŅAx A]EۂH = H=wH= HHHxH=wLxHH=1 1HHHDžLHHAx AkHtHxx zmL-6 H=% IULH HH~u=wH H5< HGHHPzILMyH x lL-R6 H=% LxIUL`LxHH Hu=wH LxH52 HGHHxLxHpHp;xH x plHJ E1AI9BwHPL0LLxHDžHLxL0HIvH/ HP =wHpLLL)L0HLH?LxH HJ4L0LxH@MtAx AHyHpHDžx 9yAEx AE?yAx AGyH@HDžHu=wL@H1HH=~6 HDžLHAx AqxHwHDžHDžHDžHDž LxLH01HH1IH٭H;AH;DYL;=N YLAądAx AwEԶHHH=wHHIHbIHLh LH1HHH=4 HLXHAx AjwHDžAx A]wHHJH58 HIHH莜ÅAx AztbP1HAjHE11 gIXZMMH8 HLAx AH=1 ZHIHH5}+ HAIHtAEx AE_1H E1HI9^HHLH?H)LHHH H4 DLHAx AHH 1ɺHH'QIHJHHH HAx A.x H =w1HH HHHXHHH]3 H 2 HP =wHH=2 HHHH Ix NMHH=/ XIH;H54 H?HIHAx AH AE1I9F3E1ɿLLLGHHH1 HP =wHHLL)HLH?H HJ4LIx rAx A'E1LMqIEH H9@H; @@˹I}HH9IU H=wI](=wAEx AEHHHHE1L=wH=*1 HE11HHL9HInMGAx A؊A$=wA$H=1 1HHLHLIHM&A$x A$H5D0 H=IHWHHIHAEx AEHH@L`pMI|$Ha HLJIH׋HHHAT$AAx AEAx AH5g/ H3<IHˌHHIHZAx A:HLx ōHHLHHfH=, HLeLMU@H`HKLE11E1HDžE1HDžHDž0HDžHE1HDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPHDž`ƅDžkƅE11E1HDžE1E1LHDž01LHDžHLHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPHDž`iHDžE11LHDž0E1E1E1HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPHDž`ƅDžkL`IH=W HLHHHPdLHPHH`dHDžE1HDž0E1(HDžLAx A!HDž鑮H`I鮯xuHLpLpE11E1YH=f HLHHHPsHPHH`7H LH5pHE1H81E1E1E1LLL0HLHLLLLLLL@LLpL LPƅDžk8DPHH=wL-& H=4 IULIHF=wAIBLLLpH5" HHNLpH H HHOAx A/HHfHHH , HHIHNH( IE wHPHH IE(wH LH=" HHHHHHHHx +EH x 4HDžAEx AER4HDžHrHƅHDžHDžHDžHDžHDžHDžHDžHDž@HDžpHDž HDžLE1HDž0HDžHE1HDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPƅDžk VHPH`麫D1H= HLOLM%*H`H;E11E1E1HE11E1L0E1LHLLLLLLL@LLpL LPL`ƅDžd-DHTzINMFfHn`=wA=wAAx A0,LHH0LP)15H0LPHINjHLP醪HDžE11LHDž0E1LE1HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPHDž`ƅDžkL`H LH5xjH81THDžE11E1HDž0HDžHE1HDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPƅDžd鱾H`HDžE11LHDž0E1LE1HDžHE1HDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDž`ƅDžkHPLuAƅ2AEx AEEH1L-} H=. IULIH2=wAIBLLLPH5 HH2LPILM1Ax A"A=wAH5%" 1LL0TL0HHPHtCH= 1HHLH9L0IAx A1HPx .#HDžAx A+#HDžMAL;L;U L;- H L豺AƅGBAEx AE)EMx\qfTs;IH8H`H HIH8AEx AE9L;AL;DlL;& _LLxLxATAx A9LxEL- AE=wAEHL E1LHHLH,HPHAEx AEHPULP11Lf3Ax A11LxE1HƅHDž~HDž01E1E1HDžHE1HDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPDLLP7LPZHDžE11LHDž0E1E1HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPHDž`ƅDždLpLhA=wAAE=wAEH`x &HHLLHC*HIAALʾH LH5~`H81Z1HLE11E1HDžE1HDž0|DMuM}AfInP=wAA=wAAEx AE4%HL)W)H`HAALڽ閟HUdH+%(dplH=C-1LH褽L HHLpH`E/$HbLULH8莽H= HLXLM3H`HE1E1E11LE11E1L0E1LHLLLLLLL@LLpL LPE1H`ƅDž_3LxZx2LLP@LPHDžE1E1E1HDž01E1E1HDžHE1HDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPƅDž_zkL0H`LLpILpƅE11E1HDžE1LDžLHDž0LHDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž LpHcH; "LH`H\H`H; AH; { DH; L`LAAx AfSE111E1ƅE11E1HE1E1LDždH0HHHHHHHHH@HHpH HPL`A{07LpLhA=wAAE=wAEH`x !HHLLH#HIA؞A̞L脸鿞HDžE11LHDž0E1E1E1HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPƅDžpHDžE11LHDž0E1E1E1HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPHDž`ƅDž_LLP۶LP鞵LLPLP阵L謶鞵LpLP蔶LpLPDLpCLcNDž]ILLP=LPLLP"LPHLH@LL )۵H@foLL LxE1E11HDžE1LE1HDž0HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPƅDžt I遧LLPѴLPQE1E1E1HE11E1ƅLHPLHDžDžHDž0HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž H酴L޳`LLPʳLPyHx貳鹳L襳H蘳޳H=o HL許LMV胭LH H){1ƅE1E1H1E11H01LE1HHHHHHHHH@HHpH HDžHP鼾fL訲LPML蔲L臲 LE1E11HDžE1HDž0LxE1HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPƅDžvԽ豫IEL蘱锳H; 'LߺLx"LxHHP[RHPH;5 H;5 H;5I HPLxHPLxAŋx qUEGAxAuLְE齣HE1E1E1HDž1LHDž0LHDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HPƅDžIL߯AxAuLïH=*IHDžHLPۯH=l HL襭LLPMѬyLLPHH |HE11E1L0E1E11LHLLLLLLL@LLpL LHHPƅDž鯺DLE1E11HDžHDž0LnHakGIx -Dž^HH(,L8LxE1E11ƅE1LHDžDžvHDž0餡H; )GLPXI`HE1E11HDžLHDž0LHDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HPƅDžLPI}ĦIxt6AxAtDž]"LLp藬LpHLpLpHH@RUHHPHуH)H-HHHh.H>.H胥HH THCSHSHуH)H,HHH-Hu.HHpHpq.UU+H w H HHp's.H輪IHqHpH0蝪L0HI6tHHL0ԥL0HHIsAx AQLLߺL L0iL0L HIfAx AXAx AX1LHlAŅ:eAx AXE\-H RYffH*pHH^dHL5( LhHHLHHQH HH0ɫH@H eH@H0HHWH H@H@H,gHHL5p HLHQH HH0LHHaVH@H0HHVH HHCVH H9GTgHGHOH0=wH0=wx VH0HHH HDžH IƋx oVH0x dVMUAx ATHpH0HH+LqHHHHNHHpHHHHL97HpL1E117KLITILKTIL9HuH0ڢH@L5 H@LMgH=0g1LH@AIH 6MSfH@x SH x SHHHIHeHH1E1E1PH1jH$4LHHHpZYAx AS1HHp=wLpHH=] 1HH1HLH*LH [Ax A%9H [rHx 8H H1HHPH5= H2HH;5 H ш51E1LE1L01E1E1HHLHLDžHHHHH@HpH :EL- AE=wAEH HLHHDžH,HPAEx AE1HPJLP11LALx AS1LxE1E1ƅHDžDžwLOA鈖H:MCMkAL=wAAE=wAEAx A1ME1¡LޣqHpE11E1HDžE1LHDž0LHDžHHDžHDžHDžHDžHDžHDžHDž@HDžH HDžpƅDž"HDžE11LHDž0E1LE1HDžHE1HDžHDžHDžHDžHDžHDžHDž@HDžHDž ƅDžlLZmLML`+茢H= HLVLMNLx*LxHHPo11E1LxHE1E11L0E1E1LHHHHHHHHH@HHpH HPƅDžyvH`EJ?H(2LHDžLLPLP$HHLpHPǠLpHPfHpE11E1HDžE1LHDž0LHDžHHDžHDžHDžHDžHDžHDžHDž@HDžH HDžpƅDž L)HL0L0LӟəLPI#H= HLϝLM誙IHbHDž1E1E1HDž0E1HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPƅDžm鋕辘INL讞L衞HDžE11E1HDž0LxE1HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPƅDžy颔LH0LP)轝foH0LPHDžE11LHDž0E1LHDžHLHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž ƅDžIJMrfHnP=wA=wAAx A /HLH0)H0IŋpeHRXLxE1E11HDžE1E1E1HDž0HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž ƅDžy鮧H`E HhL ƅE11E1HDžE1LDžLHDž0HDžHHDžHDžHDžHDžHDžHDžHDž@HDž龦HDžE11LHDž0E1LE1HDžHE1HDžHDžHDžHDžHDžHDžHDž@HDžƅDžHDžE11E1HDž0hHxړLPI黆HLH踙LH.HDžE11E1HDž0E1E1HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž HDžPƅDžm L)`ߘfo`Lʘ$HH0LH诘H0LHÜHLH荘LHĜLLHrLHǜH^=LQE1ffA.GADEE1E1ƅE11E1HDžE1LDžHDž0HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž mL@ƅE11E1HDžLDžHDž0HDžHHDžHDžHDžHDžHDžHDžHDž@ӎMnIVAE=wAE=wAx AI1ЅHDžE11LHDž0E1E1HDžHHDžHDžHDžHDžHDžHDžƅDž{LHH@E111E1LE1H0L蹁СL謁E遦H`=wL`rLupH; +#L~HHH/HHH;AH;D/H;52 }/HHxHHAŋx v/E/11L1DžE1E1LH0LHH鰌HLp藀LpHE1E1$H`tg11LxE1ƅ1E1HDž{H0UH H LxME11HDžE1E1HDž0HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpHDž ƅDž{vLPrLLH<LHÆSHHpwHQHHxLpH L LHLHE11E1HDžE1LHDž0HDžHHDžHDžHDžHDžHDžHDžHDž@HDžHDžpƅDž/H`~V~fL~lCSHH HHp;HFVHH HHBRHH HHHL}LLLHt}LHކL`}LL0L}L0遅CSHH HpBwH,}+H H=wHt HH= yIH/H5 HH0L0HHIJ0Ax A0H= L0L0HI.H5= HLH0L0LHH H$.Ax AB0HР E1I9B-HLE1LH0LH|LHI,H] H0HP =wHLH H)H?L0H HHLH4HXrLHYhH L0x +AE1Lx A+Ax A+HE1LH+H@H= H@YIH?H5f HH0L0HHI/?f )H*Y)f/H,HHH HHH HHH HHH HHH HH H HLhAx A*LL0pL0HIqZHL0H pL0L HIYH= LL (L LHH0 YL0H5 LLLLLHH HVZAx A$ZH=ϝ 1ɸI9xZHƺLHH)H?HH H HLLH4L0HLHIeLL0Ax AYAEx AE:ZH x W1HAx AWE1LMVH14L;5 Dž QWHHHLi- WHHHH0H~VAx A WLHH5 I IHH;u MUIH5? Z HHHAH H9A@LALqA=wAA=wAx PAL1HHHLH?HLE1H H4HH LHIcH x @1HME@Ax Aj&rHHpHH+H9HMMMHkpH@IHL0LHHH HHHx11E1HLLH0tH0LHL!LI4HHtH9HL!I4HHuHH HHH9pxLMMMՃ}>H]pH5 1LIAEx AE>MpSAALtLH!IH:tHL!IH:uHH@L-t AE=wAEH# E1LHHLHIAEx AEMt11LhAx A1E1LxƅHDž|H H5E1E1H8 u11LxH1E11H0E1E1LHHHHHHHHH@HHpH ƅDžrnHxDs^L7srHPLE1s1LxE1ƅHDžw-HrxHrx1ƅE11H1E1E1H01E1HHHHHHHHH@HHpH HPDž` iHLP6rLPDaL"raHraLLPrLPLqhbLqHbLqa1E1ƅE1H01E1E1LL1E1HHHHHHHHH@HHpH DžK}H9qyHLx%qLxHLx qLxu1ƅE11H1E1E1H01E1HHHHHHHHH@HHpH HPDže:g1ƅE1E1H11E1H01LE1HHHHHHHHH@HHpDž{INMnfHn =wAE=wAEAx AM@HLHp)HpIË HLpVoLp111LpHE11E1H0LE1HHHHHHHHH@HHpH ƅDžzHphI骍nH= HLlHHpHJhH H91E1E1E1LE1E1E1LL1E1L0E1HHHHHHHHH@HH ƅDžy1LxE1E1ƅHDž}_aLmotLH0) |mfo H0L`mLSmAADžbHgIsHmHplLHHlHHI"mH= HLjLM[fHPH911E1E1H0E1H11ƅHHHHHHHHH@HHpH HPDžnbHg LH5 LHE1E1H81e1E1L0LE1E1HHHLHHHHHHH@HHpH HPƅDžkaLFkH LE1E1(k111ƅE1E11HLE1DžH0HHHHHHHHH@HHpH v111ƅHE1E11H0LE1HHHHHHHHH@HHpH Dž v111ƅHE1E1E1H01E1E1HHLHHHHHH@HHpH Džu11ƅE1HL1E1H0L`Džn#]E1E1LxE1ƅLE11LDžtL0\[iH= HL%gHH H^bLHF1ƅE1E1H1E11H01E1E1HHHHHHHH@HHpHDžHx1HBtLx|hH= HLFfHLxHH bLLxHsC1ƅE1E1H11E1H01E1HHHHHHHH@HHpHDžHx1HXsLH^111LxHE1E11H0E1E1E1HHLHHHHHHH@HHpH ƅDžurH`fxH;̊ ʸH5 LOdI!|H`kf鉬1E1E1HpHE11E1LpƅHx1LH1H01DžHHHHHHHHH@HHpH qMBMjAL=wAAE=wAEALx ABME1H111HE1E11Hx1E1L H0LHHHHHHHH@HH HƅDžpH ^LxHpH111HE1E11Hx1E1L H0LHHHHHHHH@HHpH HƅDžpH ^I顅H111HE1E1E1Hx11E1H0LE1HHHHHHHHpH HƅDžsoLacRH@Pc~LCcLx餆HLx(cLx鬆LLx cLx馆Lb鬆HHb酈Lb閈E1E1E1ƅLE11E1L0LE1LHLLLLLLL@LLpL E1DžnXIMfInMufHnfl=wA=wAAEx AE)HLH0),H0HPHR RHaQE1E11E1LL0 [ImQ1E11H1ƅE1HHHHHHHHH@HHpH HPDžnWIKICfInHPfHnfl=wHP=wAx A](HPHH0)H0IƋ PPH]`OE111LE1H0Hx/ZLPIoOH0ZIOE11E1ƅHE11E1L0LE1LHL`LLLLLLL@LLpL LPDžn#VLk_PH^_PLQ_PLD_^:YIME1LE11L0LE1E1Džk1E1LxE1H011E1LE1E1LHHHHHHHHH@HHpH ƅDžtj1H01L0E11DžH0HHUE1ffA.FADEddd^H= HL.\HH0HLXIH.%1ƅE11H1E1DžnH0Ü1M1E1H1H0L]]E1E1E1ƅL0E11E1LHLE1LLLLLLL@LLpL LPE1LDždSHHH9PKOHuHPH;€ 2OpL-x AE=wAE1HLHHHb HIAEx AE LM 11LLp[ALpA111ƅHE1E1E1H01E1E1DžHHHHHHHHH@HHpH gH}[Lx{\H,H? E1E11ME11DžE1HHSg0ULL0HH6cMBMjAL=wAAE=wAEALx AbME1c1E1E11H0E1MDžf1LxE1E1H11E1H01E1E1HHLHHHHHHH@HHpH ƅDžw"fSLHI`aLH4ZH=Š HLWLLHMaSLHH1H11E1E1H01E1MHHDže11E1E1H01E1MHHDžEeL3YH0?HYLLYTLYLL L0XL L0RWHXYJH`X~JLXH LE1E1X111ƅE1E11HLE1DžH0HHHHHHHHH@HHpH dHmHpK RH1E1E1ƅLE11E1L0LE1HHLHLDžHHHHHH@HHpH Tc1L1E1H01E1LHHLDžcHpHHHH0LVL `L)PVfoP`VH= HLTLM]PHHH1LHH;z RLRHHHHHH;AH;DH;5{ }HHMHHAŋx vE\E11L1L0E1E1LHHLDža11ƅE1H1E11H0E1E1E1HHLHHHHHHH@HHpH HPDž_a1ƅE1E1H11E1H01LE1DžHHHHHHHHH@HHpH y`ML_T^LRTLETL8TLp1E1E1ƅLE1E1E1L01E1E1DžHHHHHHHHH@HHpH _LSHSpLSD1H H@kSH (111ƅHE1E11H0E1E1LHHLHHHHHH@HpDž^Hx LH8MH@LLx 111ƅHE1E11DžE1E1H0HHHHHHHH@HpH :^= H RH EHQ鄩Q钩H@=XNLL0QL0%LQ+LQMHHEWWHHkQy1L1E1H0LE1DžL;]H5 H=1H H111ƅHE11E1DžLE1H0LHHLHHHHHHH@HHpH q\PH=C HL|NLML0PJL0HHHtKE1E11LLE1E11L0E1HHDž[LOH+u LE1E1H5LH81ZI1L1H1LE1H0Dž[LnOE1LE11L0E1E1Dž\[H6OL0LL0OL0LOLL0NL01E1E1H 1L E11H1E1LH01ƅDžHHHHHHHH@HHpH \ZMjMrAE=wAEA=wAALxAt M1LN1E1ƅE1H01E11LLE1HHHHHHHH@HHpDžY1ƅE11H1E1LH01LHHHHHHHH@HHpH DžY1E1ƅE1H011LLLE1HHLHHHHHH@HHpH DžxXLL0_LL0111ƅHE11E1H0LHHLHHHHHH@HHpH DžWLL0KL0H11LHH1E1HH01E1HHE1E1H0Lx1E1Hx1LHHLHHpHHƅDžWLK;vHL11HH1E1HH01E1HHE1E1H01LH1LxLpLHHHHxHƅDždVLRJuH11LxHH1MHE1E1E1HH1H0LH0Hx1LHHHHpHƅDžULI4uH1LxMHE111HME1E1LH01E1HHLH0Hx1LHHHHpHƅDžULItHLxME1HH1E1H0E1E1LHHƅLHx1H1H01DžHHHHHpnTL\HqtHE1E11H11E1HLp1MLLxE1E1HHH0LL0E1HHLHHxHƅDžSHPHuk IMH H5LxH0H81L*AH111LE1E1H1E1E1HƅHLHxDžH0HHHHHpSHML1H11E1HLx1HH0E1E1E1H1LHHpHH0Hx1HHHƅDžiRLWFrHLpM1H11E1HL1HHHE1E1E1H1LxHHH0ƅHx1LH0HDžQHE.r111ƅHE1E11H0E1E1LHHLHHHHHH@HpH Dž-Q1E11ƅL0E11E1HLE1HHLDžHHHHHH@HHpH PHi LE1E1H8?1E11L0E11E1HLHHLHHHHHHpH ƅDžP1ƅ1E1H1LH01LHHHHHHHH@HHpH DžOHh LH5>H81=LH1E11ƅL0E1E1E1H1E1E1HHLHHHHHHH@HHpDžNH.h LH5E1MH81d<1E1E1HL1E1L0DžNHHE8I-IHHWBy1ƅE1E1H11E1H01E1LHHLHHHHHHpH DžMHLHLpALHLpfHH0H?)ԬI 1ƅE11H1E1E1H01LHHLHHHHHH@HpH DžM;HpHH@wH@B111ƅHE1E11H0E1E1LHHLHHHHHH@HpDžNLH@H@1E1E1ƅLE1E11L0E1E1LHHLDžHHHHHH@HpH KLL@1L>H A H@1E11ƅL0E1E11HE1E1LHHLHHHHHH@HpDžJHgb H5hH8@H@H@%1ƅE1E1H11E1H01E1LHHLHHHHHH@Dž8J1ƅE1E1H11E1H01E1LHHLDžHHHHHH@HpH I111ƅHE1E11H0E1E1LHHLHHHHHH@DžHW LH5WE1E1H81-,11ƅH11E1H0HHHHHHHHH@HHpH Džn(HZW LE1E1H5LH81+111LxE11E1HLH0LHHHHHHHHH@HHpH ƅDžyc=1ƅE1E1H1E11H01LE1HHE1HHHHHHH@HHpDž!1111H1LHHLE1H0LHHLHHHHHHH@HHpƅDž2E11E1LxHE11E1L0E1E1LHLLLLLLL@LLpL ƅDž~.HK LE1E1H5E1H81111HE11H0HHHHHHHHH@HHpH HPƅDž_HK LH5xLLE1H81C1ƅ1H1E1E1H0LHHLHLHHHHHH@HHpHDžHP1Hp1HMLHE11E1HH11LHHLxH0HHx1LL0E1HHƅDžs0HI LH5"LpLxH81LpLxQH@HLxLH0=wH@E1HƅHHLHxLLp(HHE11LxME11Hx1H0E1L0LE1HLƅLHDžHHHHHp3/1E11ƅL0E11E1HLE1HHLHHHHHH@H Dž.HH LH5fLxH81;Lx鶹LLxp"LxI1HLHH1H01H0E1E11HHHH1LxLHpHLHxHƅDž-M~I^A=wA=wAHx A IE1JDžE1111E1H0E1HHHHHHHH@HpH E1L1ƅ1E1H01E1HHHHHHHH@HpH Dž}E1ƅLDžE1LL0Lw HL0c L0LO *1ƅE1E1H1L1HHHHHHHH@HpH Dž+1ƅE1E1H1L1H0HHHHHHHH@HpH Džo+111ƅHE1E11H0LE1HHHHHHHH@HpH Dž*H0LL齥11ƅE1H1HHHHHHHH@HpDžLeL0IHMp=wA=wAAxAt%M1$LL0 L0髥LHL0L0H1LHE1H0HpLLxHH11E1ƅE11E1HHLpH0Hx1LHDžHHHHHp@)L.FLHMH0LpLxLMHFLHH0MLLxLkHHHE1E1E1Dž1E1HHHx1H1H01HHHHHp](LIHAEx AEIGLLAHIHGLAHHeLA־HQAFAELEL11LxH1ME1HME1E1H0L1HHx1LH0HHHHpHƅDž1'LLLxIMHME11HH0ƅE1LLDžHx1H1H01HHHHHp&LHMMH0LxLAAEx AELxVLLxHHE11E1L0E11E1HE1HƅHx1HDžHHHHHp%LHMIE1H0LxL.LLxLx(LLE1E1L辚E1E1E1HHL1L0LE1HHƅLHx1DžHHHHHp$LpLx1LxLpHHE1E1E1Dž1E1HHHx1H1H01HHHHHpd$H; H5LpLxH81LpLxYIUH =w IUHZH= B BLBH LxH0LHHHE1E1E1LE11E1Hx1LHL0LE1ƅDžHHHHHpN#H<>H/?L"?LML1H111HE1E11HpHE1E1LxH0HHx1LH0HHHƅDž"LHH0LLxHpLE1E1H111HME1E1LH01E1L0LxE1LHHHLHxHƅDž!L9f.DUfHp fHnHhHAUIATSHXdL%(LEI)E~  HEfl)EfHn)EHrLIHMbI0 MtIfH=wHUHHH]PJ4HHU6Y^dH}HHEHH}H 9 HOMLMA$wA$H"d HH5o[ AQLjRhj@IuPjRHH=J9 |s IA$HPMx A$GIH]I<$Htx VII9uHEdH+%(HeL[A\A]]@IMIH j1H}7 HH]L H5OH8AP1IH9 XZIH]I<$Htx IL9uHH=E10DIHV=wHUHV=wHUHwHEHEHH}HH 7 H]ffDHp vDHp =wHUf.wHMHDHp wHE^IHV=wHUHV=wHUDLH1p wH}HEHH 6 wHMHH] fDH .xA$uLSH H= UfH@l IfHnHhHSHhdL%(LUI)E~ HEfl)EfHn)EH^LIHMNI,MtIH=wHUHH]ML]HvHUHPJ4LU_AXH}LUL]H}uM~Sff.II:J<uH3 HH H5{L AHH8AR1 Y^HH]H8Htx HH9uHblH=,_1HEdH+%(=H]HÐIIIZLA=wAHULMHH}HH 24 H]MA=wAHIsHU AQLjhj@RHjH=k4 LUm LUHHPAHx A|HH]H8Ht xtHHH9ufHV=wHUHV=wHUHMHEHMHE롐HEHENfDMH AL H1 HH]HDzH5nyH8AR1bXZIHV=wHUHV=wHUDHk =w LMH}HUHH O2 =wHMHH]fLHM HMoH AL LMHUH}H 1 H 1 =wHMmfHj =wHUBfxAt%H/H=HM(HMLHMC HM UfHx^ IfHnH0 HSHhdL%(LUI)E~ HEfl)EfHn)EH\LIHMLIMtIH=wHUHH]ML]HۯHUHPJ4LUz_AXH}LUL]6I~&ff.IIJ<uH. HH H5vL AHGH8AR1Y^HH]H8Htx HH9uHbH=T_1HEdH+%(H]HÐIt IHHV=wHUHV=wLHUA=w AHUH}LMHH / H]MA=wAHIsHP AQLjhjx RHjH=^/ LUh LUHHPAHx AWHH]H8Ht xtHH9uDHMHE; HMHE֐HE' HEnfDIH AH, HHXL H5tH]H8AR1XZf.IuHV=wHUHV=wHUfH %AmfDH y- =wHMHH]BLMHUH}H E- 'LHMHMH - =wHMf.xAt%HH=HM舷HM8LHMHMUfHY IfHnHATSH`dL%(LUI)EH0 H)EfHn~ HEflHE)EfHnfl)EHt.LIHM~"IxHiJcH>It IPHV=wHUHV=wHNHU=wLHMA=w AHMH}LMHH+ H]LeMLEA=wAHHM IsRHH=+ jAPAjx QLj LUe LUHHPAHxADH;Htx HL9uHEdH+%(HeH[A\]IH ,AH2) HHL WH5qH]H8AR1LeXZff.H;Htx gHI9uHɨ!H=~ƴ1AHV=wHUHV=wHUHV=wHUH=wHUHL]H]LeHبMHLPJ4LUs_AX'H}LUL]I~,ff.ff.IIJ<uH' HH H5oL AH<H8AR1Y^HMHMfDfDH EAfDH( =wHUHH]LeLMHMH}H]( LHM4HM[H9( =wHUf.xAt%HgH=|HM訲HMLHMHM8UfH(^ IfHnHATSH@dL%(LUI)E~ HEfl)EHLIHMIeIEMHL]H]LeHMHLPJ4LU>q_AXH}LUL]?M"J|NIIt J|8LMH}H& `It IHV=wHULA=wAH}LMH>Hg& H]LeMA=wAH P Hj_ HIsRAHH=% jQPjQLPj5FS LU ` LUHHPAHx AH;Htx |HI9uHEdH+%(HeH[A\]@MH %AL bH$$ HH]HäH5kLeH8AR1XZH;Htx HL9uHͣlH=zʯ1PHV=wHUH=wHUH eAL ǫDH$ =wHUfHA# HH H5kL {AHɣH8AR1Y^ ;fDHM'HMrfDH)$ =wHUHH]LeLHMHMxAt%HH=xHMxHMLHMHMUfHXV IfHnHHSHhdL%(LUI)E~ HEfl)EfHn)EHLLIHM<IMtI H=wHUHH]ML]H-HUHPJ4LUl_AXtuH}LUL]*I~IIJ<uH,! HH H5iL JAHH8AR1Y^HH]H8Htx HH9uH H=H H]LeMA=wAH !E HT HIsRAHH= jQPjQLPj5G LU|T LUHHPAHx AH;Htx |HI9uHEdH+%(HeH[A\]@MH AL ҝH HH]HMH5f`LeH8AR1VXZH;Htx HL9uH=LH='o:1PHV=wHUH=wHUH ՗AL 7DH! =wHUfH HH H5_L AHSH8AR1gY^ fDHMHMrfDH =wHUHH]LeLHMLHMxAt%HH=mHMHMLHMHMxUfHpH IfHnHHSHhdL%(LUI)E~^ HEfl)EfHn)EH\LIHMLIMtIH=wHUHH]ML]H͖HUHPJ4LU]a_AXH}LUL]6I~&ff.IIJ<uH HH eH5b]L AH9H8AR1BY^HH]H8Htx HH9uH"H=H H]LeMA=wAH Q< H K HIsRAHH= jQPjQLPj5&< LUK LUHHPAHx AH;Htx |HI9uHEdH+%(HeH[A\]@MH ŏAL H HH]HH5WLeH8AR1XZH;Htx HL9uHm H=fj1PHV=wHUH=wHUH AL gDHQ =wHUfH HH H5VL AHH8AR1Y^ fDHMHMrfDH =wHUHH]LeLHM|HMxAt%HQH=ieHMHMLHM3HMUfHE IfHnHATSH@dL%(LUI)E~ HEfl)EHLIHMIeIEMHL]H]LeH0MHLPJ4LUX_AXH}LUL]?M"J|NIIt J|8LMH}H1 `It IHV=wHULA=wAH}LMH>H H]LeMA=wAH !8 HF HIsRAHH= jQPjQLPj57 LU|G LUHHPAHx AH;Htx |HI9uHEdH+%(HeH[A\]@MH AL ҐH HH]HhH5fSLeH8AR1VXZH;Htx HL9uH=VH=b:1PHV=wHUH=wHUH ՊAL 7DH! =wHUfH HH H5RL AHnH8AR1gY^ fDHMHMrfDH =wHUHH]LeLHMLHMxAt%HH=iaHMHMLHMHMxUfHhA IfHnHATSH@dL%(LUI)E~j HEfl)EHLIHMIeIEMHL]H]LeHMHLPJ4LU~T_AXH}LUL]?M"J|NIIt J|8LMH}H `It IHV=wHULA=wAH}LMH>H H]LeMA=wAH 3 HB HIsRAHH=9 jQPjQLPj53 LULC LUHHPAHx AH;Htx |HI9uHEdH+%(HeH[A\]@MH eAL Hd HH]H@H56OLeH8AR1&XZH;Htx HL9uH H=^ 1PHV=wHUH=wHUH AL DH =wHUfH HH ZH5WNL AHFH8AR17Y^ {fDHMgHMrfDHi =wHUHH]LeLHMHMxAt%H H=i]HM踑HMLHMHMHUfH< fHnHhHAUIATSHXdL%(LEI)E~ HEfl)EfHn)EHzLIHMjI@MtInH=wHUHHąH]PJ4HHU6PY^lH}XHUHHuH  H_MLMA$=wA$HHw' H=  AQLjhj@RHIujz? IA$HPMx A$UIH]I<$Htx dII9uHEdH+%(HeL[A\A]]f.IM'IH b1Hu HH]L H5GKH8AP1IH?1XZIH]I<$Htx IL9uH H=ZE1 *DIHV=wHUHV=wHUHwHUHEHHuH H  H]f.fDH< vDH< =wHUf.=wHMHfDH< =wHUOfIHV=wHUHV=wHUDLH!< =wHuHUHH  =wHMHH]D{fDH xA$uLCHn H=XUfH08 fHnHhHAUIATSHXdL%(LEI)E~l HEfl)EfHn)EHzLIHMjI@MtInH=wHUHH,H]PJ4HHUKY^lH}XHUHHuH @ H_MLMA$=wA$HH" H=X AQLjhj@RHIuj: IA$HPMx A$UIH]I<$Htx dII9uHEdH+%(HeL[A\A]]f.IM'IH ~1H HH]L H5FH8AP1IHXZIH]I<$Htx IL9uHp~s H=rVE1j*DIHV=wHUHV=wHUHwHUHEHHuH H ; H]f. fDHA8 vDH)8 =wHUf.=wHMHfDH7 =wHUOfIHV=wHUHV=wHUDL8H7 =wHuHUHH  =wHMHH]DfDH ~|xA$uLH^| H=`T[UfH3 fHnHhHAUIATSHXdL%(LEI)E~ HEfl)EfHn)EHzLIHMjI@MtInH=wHUHH|H]PJ4HHUFY^lH}XHUHHuH H_MLMA$=wA$HH7 H=P AQLjhj@RHIuj:6 IA$HPMx A$UIH]I<$Htx dII9uHEdH+%(HeL[A\A]]f.IM'IH "z1H5 HH]L ]H5BH8AP1IH{XZIH]I<$Htx IL9uHy H=RE1ʅ*DIHV=wHUHV=wHUHwHUHEHHuH H H]f.kfDH3 vDH3 =wHUf.=wHMHfDHQ3 =wHUOfIHV=wHUHV=wHUDLH2 =wHuHUHH { =wHMHH]D;fDH wxA$uLHw8 H=O軃aUfHP/ fHnHHAUIATSHXdL%(LEI)E~\ HEfl)EfHn)EHzLIHMjI@MtInH=wHUHHwH]PJ4HHUVBY^lH}XHUHHuH H_MLMA$=wA$HH H= AQLjhjRHIuj1 IA$HPMx A$UIH]I<$Htx dII9uHEdH+%(HeL[A\A]]f.IM'IH u1H HH]L zH5g=H8AP1IHwvQXZIH]I<$Htx IL9uH0u= H=ME1**DIHV=wHUHV=wHUHwHUHEHHuH H H]f.fDH/ vDH. =wHUf.=wHMHfDH. =wHUOfIHV=wHUHV=wHUDLHA. =wHuHUHH =wHMHH]DfDH >sxA$uLcHs H=KUfH* IfHnHAUATSHHdL%(LEI)E~ HEfl)EHLIHMIsISMHLUH]LeH~sJ4HLP=Y^LMLUMgHuH H=wHUHfIIMMH qH HH]L vH59LeH8AP1IHrXZH;Htx HL9uHnq H=Jk}1@HV=wHULA=wAHuLMH~Hg H]LeMAE=wAEH  Hh+ HARH= HIrjQPjQLPj5( , HAEHPHBx AEH;Ht xt4HL9uHEdH+%(DHeH[A\A]]f.HMwHM뽐HV=wHUH=wHUL * A=fDH o1fL a* A=wALMyLHMHMH =wHUHH]LehfDx AEt%HFo H=GHM?{HMLHMZHM@ff.UfHP& IfHnHhHSHhdL%(LUI)E~ HEfl)EfHn)EHNLIHM>IMtIH=wHUHH]ML]HhoHUHPJ4LU9_AXtwH}LUL],I~fIIJ<uH HH mH55L sAHnH8AR1Y^HH]H8Htx HH9uHm H=DFy1HEdH+%(H]HÐIt IHHV=wHUHV=wLHUA=w AHUH}LMHH = H]MA=wAHIsH AQLjhj@RHjH=n LU' LUHHPAHx AWHH]H8Ht xtHH9uDHMHE[HMHE֐HEGHEnfDIH kAH HHlL qH53H]H8AR1XZf.IuHV=wHUHV=wHUfH EkAmfDH =wHMHH]BLMHUH}H e 'LHM4HMH 9 =wHMf.xAt%Hj2 H=qCHMvHM8LHMHM8UfHX fHnHhHAWAVIAUATSHdL%(LEI)E~3 HEfl)EfHn)EHLIHMIoIMtIH=wHUHHMHjHhJ4HUP5Y^H}HUHnLeM9HEH H`=w H HEHUHHHHEHX0Iŋx ZML=Z H= IWLoHH=wHCH5 HHHINjM7x 0LLhHHlAx A1H; H; u H; Dx EI=wIvH`MHH= S# IƋM=xffDAEx AEsHhLeH;Htx HL9ufIMIH 2g1HE HL qlH5/H8AP1IHChXHEZHhHhLeH;Htx HI9uHf_H=?E1rHEdH+%(HeL[A\A]A^A_]IHV=wHUHV=wHUH=wHUHUHLeHEHhML%y A$=wA$HELeHhHY H= HSHnIH=wAIGH5 LHH HH'Ax A4HLdINjMCx L;= L;= #L;= LVÅAx AMA=wAIvH`MLH= 4 IHAAL@HADžU#A0L% A$=wA$LeH =wHUDL% A$=wA$LefDH =wHUtf.L%1 A$=wA$Le;fD'fDHIHV=wHUHV=wHUDH萿胿ZfDLpL`HHH bHbH=;E1nKH= HxLHxHH H LH5`H81\At+ff.H;bDH=';E17nbfHXfDKIaAAxAuL#lH= HxH6LxMHA]DA8EHH@AkDH蘽L舽HxLhAhL% A$=wA$H=@ H;=A IEt H;w LѳIMttHXLLuHHE$(IAxAtXA$x A$tQMt11L.Ax At;AA$xA$uLlLbLXLNLI>跺H HH5]H81̵/Aff.UfHx fHnHAWAVAUATSHH`dL4%(LuI)EHH-)EfHn~7 HhflHE)EfHnfl)EHt.LIHM~"I%HlJcH>IIIH ^AL +gH HH]H_H5&LeH8AV1蓴XZH;Htx HI9uHy^H=7E1sjHEdH+%(y HeL[A\A]A^A_]fDI<HV=wHUHV=wHUHV=wHUL>A=wAL}H}bHUH-LmH]LeMHEH=  HX=w H= HEHUHHHhHEHP$HhIƋx fMH5 H=? HVHh蟸HH=wHAHhHH5 HHPHhI‹M)x LֺLLh胵LhHHAx AMH; & H; u H; x 4H  H=> HQHHh蛷IH=wAIBLhLH5 HH.LhHHAx ASHκLHh|HhI‹Mx 8L;! L; u L; KAx AH`LA=wAH`LAUMjH= HLhL } HX` YLhHI^AAALܶHV=wHUHV=wHUHV=wHUH=wHUHH]LeML=\[J4HLAW%AYAZeH}JH}H}\M~-@ff.ff.IIrJ<uH HAL aH xYLH5r!H8AV1f_AXfD裵fDH`H=wH HH MH`AUAjH= RHPjRPj5 HXHhI HhIHPM x aAxAfff.H;Htx HI9ufHHh葬Hh EA_wHojf.LHhQHhDHωh2hOfD fDL- AE=wAELmH]LefDH! =w L}HUfDL- AE=wAELm|fDL}HULmHLhYLhcDL@[MH VAL \ fHa =wHUfL- AE=wAELmfDL- AE=wAELmfDH[V^H=}/E1Ub˲HhH=U Hx葰HxHlHA_;LLhLhAeAx At%DDHUE1H=.afLرfD@軫HhI@A_H蘱LHh聱HhDHLhaLhL׉hGh臱HhH= HxMLxM(HAe@A``تLhHHAe0L-q AE=wAEH= H;= IFt H; L1IMHPLL}HHEAx AAUx AUtgHt'1H1Hh$"Hhxt1AjAExAEuL輯AfH觯LHh薯Hh~LHh{HhRLIH HhH5QH81H HhH5PH81̨蒭fUfH fHnHAWAVAUATSHHXdL$%(LeI)EHH)EfHn~ HhflHE)EfHnfl)EHt.LIHM~"I H`JcH>It I HV=wHUHV=wLnHUAE=wAEL6LmA=w ALmHELuL}H]HHH3 HE HPH< L(hE111AHLAHhH =wƒ% H L(hE111AHLAIH=wƒ H L(L`hE111HPAH`H`H=wƒn HhAD$;F L5 H=b IVLƫIH =wAIBLPLH5 HHLPIAMx AL-/ H= IULDHH=wHAHPHH5 HHHPIM x HP I9E HhHuLLeHHEHEMHAx AHH I9FoHuLHMMHHPHEBHPI‹x AEx AEM L;2 L; u L;  DAx A EH  H=F HQHHP裩IHV=wAIFH5m LHHHHAxAuLHPHPL5 H= HPIVLHPHIh=wAEIEHPLH5 HHHPIAEMxAEuLHPIHPH I9F HuLLeMH`HPHHEHEJHPIAEx AE5MH H9AHuHLMHL0HEHPLPL0IAx AAx AuML; L; E L;F 8 LLP LPAAx A_EL5 H= IVL!IHA=wAIBLPLH5 HHLPIMAx AH H=; HQHHP蘦IH=wAIFH5: LHHSHHAx AH H9A.HhHuHHHEHEH`HPHELPIAx A M H= I9ElHuLLEMHLPHELPIAx AAx AML; L;U m L; ` LLP辝LPApAx AgEHXLA=wAL  HLHXjH= AQH`jAQATjLhLXHH} LXH@HIff.Ax AD HhHh. A$x A$ MWDIH \HAHb HHIL MH51L}H8AT1H]XZff.I?Htx _II9uHG7 H=S!E1SHEdH+%(3HeL[A\A]A^A_]fDHV=wHUHV=wHUHV=wHUH=wHUHL}H]ML-HJ4LHAU_AXH}KI~+fff.ff.IIK<uH HH FH5L LALH8AT1覜Y^fDH }FA%fDH`财 H`;ALHf.QfI~zBHP'f.wQPz7Lf.WQf(zfInf/%f/Pf.PzHXH=wfInH0 葛H0HI H0jH0H PHPFH0LPHIt H HMHX5i H=z AjHPARL0jPASjPHHHX8 HXLPHPHL0IL x Ax AEAALLX詠LXDH wHEL}H]HHf.HELuLmHH}Hh4L H) wHEfHDž`E1A{ ff.HCDH=OHhtHhx Mt#A$x{E1A$twMff.H`tH`x t2fI?Ht xtIL9u D#HfDE1L|HSLLPɖLPA7A MAAL虞L茞 ҞH=c HxL蜜LxMwA HxH LH5@H81ݗA RfMVfInMnAfInfl=wAAE=wAEAx A{HL)EH`H0LPHELPH0IAALLPqH0LPnfDHDž`A| rfA} eDL(LHPHPDHdD2H= HxLHxHHXЖHXHu$H LH5]>H819HXfA AxA2H}rH:eDLLP!LPSDHLPLPDA$A$LM̛aHhTL豛H褛A$A$|LLPLXtLPLXHLXRLPLX]H`LHA H3HA /L- AE=wAEH HuLHHEHE IAEx AE>Mt11L Ax AYA 0艔0H%A }A x AH=DH=NIl@LPIL- AE=wAEH HuLHHEHEIAEx AEuMt-11L AxA ALxA L-6 AE=wAEH HuLHHEHE{IAEx AEMt-11L2 AxA ALA ђHPIx A oLH0LP萘H0LPDMUMEAfInh=wAA=wAAEx AE-LǺHLeL0LP)EUL0LPHAALH0ΗH0LPpL賗E1A Dx M< A1ɅxAtEM AADLHXAHXE1fLHPLXLXHPf.HLXLXVLLP֖LPpLLP軖LPBMFfHnMnAfInfl=wAAE=wAEAx A(HL)EH0LPLPH0IA>A2LLP H0LPAA GAt;JLLPHX躕HXLPAA yL荕L-Q AE=wAEH( HuLHHEHEIAEx AE"Mt-11LMAxA A4LA !, H=HM78HMLHMRHMdžUfH IfHnHATSH@dL%(LUI)E~y HEfl)EHLIHMIUI5MHL]H]LeH,MHLPJ4LU_AXH}LUL]/M"J|>IIt J|(HEHUWfDIt IHV=wHUHwHUHEH]LeH/MA wA H=P HIsE1H jAWQjWH=ʫ QLj5& PLU LUHHPAHx AH;Htx {HI9uHEdH+%(HeH[A\]MH )AL /Hĩ HH]H+H5LeH8AR1XZH;Htx HL9uHm) H='j51QHV=wHUH=wHUH )AL g1DHQ =wHUfH HH (H5L 1AH*H8AR1~Y^ ۄfDHMDŽHMsfDHɩ w HUH]LeLHM脄HMxAt%H'([ H=HM 4HMLHM;HM谂UfH IfHnHATSH@dL%(LUI)E~u HEfl)EHLIHMIUI5MHL]H]LeH_)MHLPJ4LU_AXH}LUL]/M"J|>IIt J|(HEHUWfDIt IHV=wHUHwHUHEH]LeH/MA wA H=@ HIsE1H jAWQjWH=R QLj5 PLU LUHHPAHx AH;Htx {HI9uHEdH+%(HeH[A\]MH %AL *H HH]H'H5LeH8AR1v{XZH;Htx HL9uH]%` H=?Z11QHV=wHUH=wHUH $AL W-DHA =wHUfHѤ HH $H5L -AH&H8AR1zY^ ˀfDHM跀HMsfDH w HUH]LeLHMtHMxAt%H$ H=HM0HMLHM+HM~UfH IfHnHATSH@dL%(LUI)E~q HEfl)EHLIHMIUI5MHL]H]LeH$MHLPJ4LU_AXH}LUL]/M"J|>IIt J|(HEHUWfDIt IHV=wHUHwHUHEH]LeH/MA wA H=0 HIsE1H jAWQjWH= QLj5 PLU LUHHPAHx AH;Htx {HI9uHEdH+%(HeH[A\]MH !AL &H HH]H"H5vLeH8AR1fwXZH;Htx HL9uHM!2H=_J-1QHV=wHUH=wHUH AL G)DH1 =wHUfH HH H5L (AH!H8AR1wvY^ |fDHM|HMsfDH w HUH]LeLHMd|HMxAt%H H=HM,HMLHM|HMzUfH8 fHnHPHAWAVAUATSHHhdL$%(LeI)E~m HEfl)EfHn)EHULIHMEIo M MtI~ H=wHUHH]HUML-m J4HAUuAXAYtxH} I~) ff.ff.IIj J<uH HLH5L #AH nH8AT1ft^_LeH;Htx  HI9uHI8 H=E1C*HEdH+%(/HeL[A\A]A^A_]fDIt IHV=wHULnAE=wAEL&LmA$=w A$LmHELeH]H`H HT L(hLE11A1HAIHx =wƒ$ H EuL(hLE11A1HAIH =wƒ H=j EAL$Mt$M}A[ *AH= ; Aff.A 'H \^f(QY'XYf/E1ff.HhH=wHH E1HhjH= AHhjXAUjpATHhH`H HhHPHI& x  AExAE ff.A$x A$[MMtAx A0LmH;Htx {HI9uDHL  H5W L  H=y LtIHmLHBxIAM7x AH H= HQHHXuIHO =wAI@LXLH5: HH LXIM Ax A*H I9F HuLLeHHEWLIx MIPH;̚ H=U  I@ H LXH¸H)APHHIt I HV=wHUHV=wLnHUAE=wAEL6LmA=w ALmHEH]LuHhH]HHH H L}L(hE111AHLAH`H =wƒ6u H L(hE111AHLAIH:=wƒ06 H_ L(LXhE111AHLXHXH=wƒY H`AD$;F\ L- H=6 IULaIH=wAIFH5d LHHaIAMx A%(f(0[IHH`Hg^IAMx AHY I9EHuLLMMHL@HEL@HAx AAx AkH H; H; q u H;   Dx O EL= H= IWL`IH=wAEIEH5ް LHHpH@H@&AExAEuL[`0ZIHEHLH \L HI Ax AH@H5 H9p*L@HuL}HHELHAx AfAEx AE6H@ H; H; ߃ $H; HH@HWH@A x 2E\L=M H= IWLb^IH=wAEIEH5+ LHHQH@H@AEx AEQL=Ͷ H=~ IWL]IH=wAEIEH5C LHHbIM*AEx AEeL-V H= IULk]IHo=wAIBL0LH5 HH[L0IMAx AHt I9GaHuLLeMH`HEHHEIAEx AE=MH I9FHuLLUMHXL0HHEHEYL0IAx AAEx AEH@MH5 H9p\L@HuL}HHELHAx AAEx AEHH; Ҁ H; GH; N :HH0TH0Ax EHPH=wHH HPH=@ jXHXjATjL c HPL`HHغ H@HPHIWxoH` H` A$x A$sM*IH AH~ HHL H5H]H8AT1yTXHEZHhLhI<$Htx SII9uHM H=E1G HEdH+%(HeL[A\A]A^A_]fHV=wHUHV=wHUHV=wHUH=wHUHHMH]ML-HhJ4HAU_AXH}HhI~DIIJ<uHL} HH %H5"L jALH8AT1SY^CYfDH A-fDHXYHX;FIF] IFHHHH)HHAHLRII IE IEHH)HuHHHHLQHH IG IGHƃH/HH)HHHLH0LQH0II fI*f/0 fH*f/ IL9 HPH=wLH0H@VH0HI* H@H0VH0HIL LH@H0tVL0H@HI HH E1HPjH=L| AXHAUjARjAWHPHH HPL0HPHIG x Ax AWAx A_AEAELVH!{ wHEH]HHHEHhGHELuLmHH,H`UzLUHz wHEfHDžXE1DžPPHH=H`Htx Mt%A$E1A$MfDHXtHXxtBLhI}Ht xtIL9u0TfHTfDE1LpTrH`TAHH@9LH@A>DžP&E1x YE1E1E1MtAx AMtAEx AE.MtAx A6MmAbAVLSIHSSH=h HxLQLxM2|MH<DžP&H@x LE1E1E1DžP&E1ҋx MAALLHRLH@HRLLHRLHDLLHRLHDžP#E1E1E1fHL0LHJRLHL08LLH!RLHDAVH)HI\AEHAGAI)Lf.HDžXDžPVfDDžPAAFAVHH IfAFAVHH IIfDAUAEHH HfAUAEHH @LQEoAGII ,@EoAGII IfL5 A=0AhJIHf H= H]GHH Ax AHuLHMHHPHE蕻HPIŋx Ax AMt!11L6AEx AEDžP'fLL@OL@-DLH@OH@zDLH@OH@GDA$\A$OLM\OqDAEAEL&OHOA$H`LLPNLPLNHLPNLPSA$HX&HHDžP2LILH$H0vHH0HDžPLH0H0IGfL5ٔ A=wA0GHB H= HHPDHPHI> x +HuLL}HHE׸IAx AAx AMt!11L}AEx AEDžPH0=GH0HEDžPDžP&AE11LHHLHHHFIL-l AE=wAEHS HuLHHEHE豷IAEx AEGMt11LhAx ADžP x DžP"x ADžP&ME1E1x HDžP#E1E1E1LKMufInM}AfInfl=wAA=wAAEx AEHL)EL@肶L@HAALL H@JL H@DžP$E1E10DIHH=Û H{AHHAME1E1E1DžP'fLhJ^LH@TJH@LH@9JH@AEE1Ʌx AEDžP!E1HID8JH=ɐ HxLHLxMCLHaH o H5mH81ICD@LHPIHPHmI>L`IDLSIDžP&AEAEL#IE1UCH@LI H@E1E1x DžP&LHLpLhfInAfInfl=wAAE=wAEH@x sHL)E诳HAALH@6HH@iL"HjLL0HL0LGLGLGLG]LGcLH0GH0/LH0GH0)LGYHvGHiGE1DžP"DžP&M1hL;GAx ADžP(E1K\GH= HxL&ELxMAHDžP*yfDžP*mLL@) Ffo L@:@H@FH=R HxLDLxMf@HH@H@E1E10DžP*'HEDžP!L@E1H@H L0LHEH LHL0HEH=| HxLCLxMuLP?LPHIH@E1Ƀ#DžP*E1E11IHAx ADžP'E1E1E1H@MыDžP*E1E11{>L0IH@H@E1Ƀ2 IOMofHn`=wAE=wAEAx A-HLLeH0)E+H0I‹XMHL0CL02Ax ADžPE1Ax ADžP0LpLhfInAfInfl=wAAE=wAEH@x HL)EFHA^ARLH0BH07+H@E1E1INfInMnfHnfl=wAE=wAEAx AHL)EHXL H0HEtH0L INj HL0AL0L- AE=wAEH HuLHHEHEIAEx AEzMt11L訳Ax AcDžP+DžP*LE1E1E1DžP&E1EDžP&E1E11.DžP-E1H) @fo qDžP*E1E11LH0) @fo H0DžPE1xDžP*E1E1>DžP(E11LL H0)A@foL H0DžP' H= IT$LR:HHO=wHAH@HH5 HHH@IMgx H^_ H0I9FHHHuLLEHLHHEHE补LHLIAx A}x _HDžhMA$=wƒW sHDž@Hd At$I|$ LM LLHHHX1AIHOL9H@Htx IFH5Ɣ LHHQIM6L@LL L@HHhHAx AHDž`x HPH5 HDžhLH IWHH8:H@HH@H8HHH@LH@HJHPH5Z LH IWHH89:HHH@H8HHLHpHHH@H;0LGHOA=wA=wx &HHuLEL0H8HEעL0H8HhIAx Ax HDžpMAx AHDžh2H8HPHH-H0Lx@LPL`L E1I0LLH0I@H0H0I8H046HENIF ED | I010fff.H(H0HH@(HA;v}LHH@HHtˀ8V H(HI8Aq HI(H0HA;v|IL9HLPH0L H80H@H51~ H@LMaH=HP /L1HPH@AIG2MH@xu H@>5Ax A A$=wPA$x A$A:AM f.4!fDI6 H tAHzX HHL H5IH]H8AT19.XZfL82f.@zsIF{IFHHl HH)HH$H8L-H8H8tH=+ @P fH= H*8( L=X L9HL5 H=mz IVL2IH =wA$ID$LhLH5' HH(IMA$x A$L5= H=y L0IVLK2L0HI=wA$ID$LhLL0H5 HHL0IMA$x A$uHBW HDžhH0I9BHHHuLLuHHEHELHzLHHpIHDž`Ax A Ax AMA$=wƒHBx At$HDžpI|$ H5Ȅ HHID$HHHPLIVHH ]3IH1H@H HHLLIHHPH5 LHIVHH 2HHH@H HHLHHH@H;0LwHOA=wA=wx cHHuLuH0HE蟛AH0x A-x OHDx +H0HPHH|LHLPL`L Lx@HHLE1@H8@LL6/JIM9uLPL HLH0*IGH5w LM^H=HP(T1HPLAI+MAx AAx AoA$=ƒMAEAEL.H u L1HXH@HHhH;HL%o H= u HDžhIT$Lx-IH=wAIBLHLH5Ѐ HHLHIL`MAxAuLLH-LHH@LHH5 HGHHLHIML% H=;t L8LHIT$L,LHL8HH=wHAL0HL8H5܁ HHHHHHL8L0ILpMx HkQ H0I9@HuLLUHL8HELuLH讗HHL8HhIAx AAx APHDžpx OMA$=wƒ5yHDžhDu;HH0H;0HHH0HH0HcI yJ5fff.ff.H@(HH(H)0HHcHL@(L;(}IL@(HH(H0H@0HH@(HH(H+0H0IHV=wHUHV=wHUDH 5AfDL=O A=wAL}H]LHXDHELuLeHH;L(*HX*L=O A=wAL}E1E1E1A L LKHDžXHhHtx sMtAx AHpHtx LDL0MtA$x A$MtAE1AMMtAEx AEjHXtHXxt M H=g HxLHxHHE1E1E1A L Lў@HH=A L@E1L LMH@IBH2g H HxH9NHXHHqH~1H;T 'HH9uHmC HJH5ʐHWH81<Ax AhL@A E1L $L۝qHuHB H5֍H8v H@x A E1L LBIMNINAfInH=wA=wAx AHHuLEL8L@HH)E҉L8HHIL@AALKL@HHL0A.HYC HH8A1 E1E1L Lp+H$C H H8LHHHHHLHHHHZHL8LHwL8LHL\zLL0HL0 HpfHuH)EHH8`H8HhILHA$ qL<H8H5 E1E1H=A lE1L TL H~A% HeHhHA L LH@1nIH@xH@ MfDA& H&A HH8AALL%b MH@HpI9AHXHHJH~1L;d HH9uHVH? H5~A IL$H81@A=AA' LL0L0pLLL A=wAIrH =wLLLIeH=Ga HxLLxM3[HL@A E1E1L ڽLLL@E1A L Ln)LHIHH0H0H@LHI-A$AAL@1E1A LLHHPWHPLHHH0=HH0H0Ax A L@E1A L0L8HHIH= LH8A* .wH`aLnH=E_ HxL~LxMYH~1HhE1E1A- L ֻLHLH H H0IpfInIHfHnH fl=w=wAx AHuHϺLuL8HH)EÂH HHHhIL8H7L8HHLIL0NH=] HxLLxL0MLPLPH1E1A- HhL ^LE1E1A- L ?LMH ]H LILA$xL3H&L0InE1A- L Lv1L!LL8H@LH) fo L8H@LHHMA E1IJMzfHnH=wA=wAAx Aq HuLLuHH)EkHHHpIċH1LIAx AMf.A$HfHuHH0)EH0-H0H>L1H0 1A$QL`fHuH)EL=HhISL LP+ LLLLLHDžhpHHpH`HhHhHpH`HH1HHLLHLIQH0H@HHH={LLLLH[ LLLL1LLH0LLLH]HLLLHL H0H0Ax AH>H;6 H;a6 kL9bHLLLH0 H0LALLx 'EEOHHtx HHtx HHtx H1HpH@xH8LHtx H Htx M?A$3A$&L@H5 LH8 H0x HhHE1A* E1E1L Lώe HeH3 H5~H8fJH@HHH9nHuH;3 \9DHi3 H5H8 9LL8HH)foL8HHLL01LIAxH0tfMnk HH2 H5}H8lH0zfDL@E1A H0H0H0H0L yL0LhM3AAu$LLHLPwLHLPA* E1E1H3 LH5H81L@LHLP HHHHuE1LPL@LhL`LpLHHH@xH8LHtx BH Htx KL`MtA$x A$SMAxAtE1E1A* LLHE1A* LP' LPLHE1QM1L@E1A L "HHI9HuL;%0 w`HI0 H5bE1A H8 E1L YLDHLLL0U LLL01H0LLLHL LLLHL] L LOaLHH) fo HHeAAwLLLHL0e LLHL0)LLLLCLLHLLt=H0xoH0AA1fIlIIH AH* HH5L H5crH]H8AT1SH`ZYHH| fDI HV =wHUHV=wHxHV=wHpHV=wHhH=wH`Hp_ Hx) H} L%wb A$=wA$H`LeH]HHHhL`H@HpHHxHHEHAE=wAEH@=wA$=wA$H5X Lm~ t"H5a Lm` j A$x A$ L=] H=K IWLIH=wAI@L0LH5T HH[L0IM^Ax AK H$) I9GHuLLmMHHE}oH Ax A H 3AEx AE L-\ H=J IUL&IH=wAIAL0LH5S HHL0IMAxAuLjH+( I9GH@HuLMHHEHE}nH0AExAEuL H0mH@xuHL-[ H=I IULIH=wAIBL@LH5W HHRL@IMAx AH L@H5U HGHHL@IMcH& I9AHI HuLL}HHEHEL@.mL@IAx AAEx AE\ML;5& L;5% 'L;5& LbAŅyAx AyEkL-lY H=H IULIHz=wAIBL@LH5YV HH"L@IMxAx A H0L@H5lS HGHH"L@IM"HL% I9@#HG HuLL}HHEHEL@kL@IAx A. AEx AE M"L;5$ L;5Z$  L;5%  LAŅ#Ax AE L-$ L9 HH@6IHJ*H=wIGHHH H5Z HGHHj#IMp#LvH@Ho$Ax AH@+%H0H56Z HGHH%IM#L HH%Ax A"HL%KE A$=wA$HrP HuLHHEHEiIA$x A$.Mt!11LGpAEx AE^6A0HDž1E1E1HDžE1HDžHDžHDžHDžHDžHDžHDžHDžHDžHV =wHUHV=wHxHV=wHpHV=wHhH=wH`HH]J4MH`L5NHHHAV lA_ZHpHxH}fH};HHI~, ff.ff.IIJ<uH HLL 5AH ՟H5gH8AT1A]A^LHfIL9t'I<$HtxuIL9uHH=mzE1腫HEdH+%(3HeL[A\A]A^A_]H5M L dADžA$x A$% EnL%B A$=wA$HR HuLHHEHEafIA$x A$Mt11LmAx AH@L AH0HDž1E1E1HDžE1E1HDžHDžHDžHDžHDžHDžHDžHDžHDžfDL 6LLLHAX =wHUf.HW =wHxL- AE=wAELp|LeDHU =wHUfHW =wHU|fHV =wHxNL- AE=wAELpIH gAH@L AH0HDž1E1E1HDžE1E1E1HDžE1HDžHDžHDžHDžHDžHDžHDžHDžMtA$x A$MtAx AMtAx A.MtAx AMtAx AHtx (HHtx HDH=uMtAx AmMtE1Ax AbMHtHx HHtx HHtx sHHtx fHHtx YHHtx LHHtx ?HHtx 2HHtx %HHtx H xtTH0xtPLHII9I<$HtxuxfDHhfDHXfDKifDH8sH(HHHHHHHLLLHL@zHL@LH@QH@DH8H(LLLLHL@LLLHL@DLLLHL@LLHL@LLHL@cLHL@fDh2L@ILL@L@?LLMA"[L%9 A$=wA$H`I HuLHHEHE]IA$x A$Mt!11LdAEx AEA$DLL@AL@D{H= 9 HXLELXMU H!H@1E1E1L E1E1AHDžHDžHDžHDžHDžHDžHDžHDžHDžHDžHDžH0LHL8r+L0I@H@1E1E1L E1E1AHDžHDžHDžHDžHDžHDžHDžHDžHDžHDžHDžH0MGfInMwAfInfl=wAA=wAAx AHL)EL0][L0H AALLzH=6 HXLLXMIH#H@A H0fH@1E1E1HDžE1A HDžHDžHDžHDžHDžHDžHDžHDžHDžHDžH0 L0I5MwMoAfIn@=wAAE=wAEAx AHL)EYH0A*AL<L/uH=5 HXL?LXMQL@L@HI$HDž1E1E1HDžE1A"HDžHDžHDžHDžHDžHDžHDžHDžHDžH4 H9HXHtJHqH~ 1f.H;T HH9uH=wLHH9THuH; uAHDž1E1E1HDžE1A"HDžHDžHDžHDžHDžHDžHDžHDžHDžkH %L@IHDž1E1E1HDžE1A"HDžHDžHDžHDžHDžHDžHDžHDžHDžIIfInMifHnfl=wAE=wAEAx A[HL)EH2 H@HEhVH@IƋ0%HL#LL?LLKH=1 HXLLXMjL@L@H-FHDž1E1E1HDžE1A#HDžHDžHDžHDžHDžHDžHDžHDžHDž'LH0H5E HGHHIMv111L*bIHAx AH0LH5D HGHHLIMl11ҾLLaLHI-Ax AlLLǺLLLLHIWAx ATAx AL;5 L;5 GM9>LTsAx AH H5C HGHH,IM111L`IHnAx A H0LH5/C HGHHLIM<111LL1`LHIAx AELLǺLLLLHIAx AEAx AML;5 L;5\  M9 LAŅAx AEOIWHBpHH@H H5. LHHuHH! H9P8LH H5A HGHHIMH111LL^LHH;Ax ARIV IFHHHH9H9wIVHH4HIFHx  H=wL1HLuH=_A HLm$IAEx AEMA=wAH LH5^@ HGHH LIM111LLLY]LLHHjAx AH1H=H? HHfInLC HE)E+LHAARHxHAx A8HH0=wL-; H=* IULIH=wAICLLH5/6 HHOLHHAx AH0H1HH=3 HUHHEH0Iŋx aHx VMH0x oH5Z? L=KEHH7H? w H? HHVHH=: 1LgIH]8Hx H5> LL0nL0HH7=wƒ7Ax AH=wHHq H9P49H1HuHHELmLHIx ;M8IPH H9t H; 9IxH7H9?IP H=wIP(H=wIP0H=wAx AE1LLLH5N5 H%Ipt"H5 2 LIRHDžH51 LH%SH2%H=wHHHH5. HGHHb=IM3=HLL05L0HI?=Ax A$H LL0L0HI<Ax A=Hx <L3IH-<H5: HLH0>lL0{;Ax A<A$=;A$Ax Af;Ax A;LML0f.1IH?1E11E1LE1E1A'HHHHHHHHHHwLL0) fo L0L@ID H0L@I/HDž1E1E1HDžE1E1A#HDžHDžHDžHDžHDžHDžHDžHDžHDžSHDž1E1E1HDžE1A#HDžHDžHDžHDžHDžHDžHDžHDžHDžIHfInMhfHnfl=wAE=wAEAx AHL)EH# H@HEGH@IƋH LHH`L)0fo0ZMA#=H IHDž1E1E1HDžA-HDžHDžHDžHDžHDžHDžHDžHDžHDžtHDž1E1E1HDžA/HDžHDžHDžHDžHDžHDžHDžHDžHDžMA-HDž1E1E1HDžE1E1E1HDžHDžHDžHDžHDžHDžHDžHDžHDžLLLLH@)foH@{L% A$=wA$HA0 HuLHHEHEDIA$x A$Mt!11L~KAEx AE A.2LLLLLH0I,MA/NLLLLLHgHRH5$QA8H814HDž1E1E1HDžA1HDžHDžHDžHDžHDžHDžHDžHDžHDžEH ILLIH3AEaHu HL4HLLvLHDž1E1E1HDžE1A1HDžHDžHDžHDžHDžHDžHDžHDžHDžH0LILH@)foH@=LHDž1E1E1ME1HDžE1HDžA1HDžHDžHDžHDžHDžHDžHDžHDžHDž1E1iLALz&1L1E1H1E1E1HA9E1HHHHHHHHHXLLALMA1HDž1E1E1HDžE1A)HDžHDžHDžHDžHDžHDžHDžHDžHDžHDž1E1,LLLmLLkH5#' L<2/^IHb2H& w H& IPLL0HH=}, 1XL0HI"Ax A"H5& LL0z`L0HH =wƒ h Ax A H=wHH H9PHE1HƺHLEH)H?LmH H4HLD>HH0HL0x pMsISH>H9t H;RI{HA!H9 IS H=wIS(H=wAx AE1LLLLoHH0IQHLLLLL4LLCL@HLvL>L%3 A$=wA$H' HuLHHEHEx<IA$x A$Mt!11L/CAEx AEqA2111E1HE1A9HHHHHHHHHHH oI7H0[LIHLH5rH81H@L AH0WL /IH+Ht! w H^! ISLL0HH=N( 1TL0HI+Ax A+H5! LL0K\L0HH=wƒfAx AH=wHHH9PL1ɸHƺLHMH)H?LmH H4L0H@:H@HӻL0Ax A{HHE1LHLLLLHJE111E1HMA:LLLLLLLLLLH LIE11E1E1LME1HDžE1HDžA/HDžHDžHDžHDžHDžHDžHDžHDž9111E1HMA:HHHHHHHHHLL11E1E1H1E1A9HHHHHHHHHE11E1LHDžE1A/HDžHDžHDžHDžHDžHDžHDžHDžHDžE11LE1E11E1E1LYLdHLE1E1H5PmA H81&H@E1LL1E1E1LLLLLLLLLH0]E111E1HE1A:LLLLLLLLLLLSHH1L01AKH1HHHHHHHHz#LH111E1HE1A:HHHHHHHHHH= HXLLXML@L@H1L01H1AKHHHHHHHHH`HgLH5jE1A"H81E1E11LE1LLLLLLLLLL@LL~iLqHL0]L0)HL0>L0SL*YHL0L0E111E1HE1E1A8LLLLLLLLL$LL0AL11E1E1H1HHHHHHHHH11E1E1HL0AY11E1E1H1HHLA1LL0-L01L01E1HE1E1An MbIRA$=wA$=wAxAt3I1HL@L0觿L@L0LH0腿H0L0HwnHx耷L0HH!1L01E1HAnvHL0AL0HH5L0HHtH`HYHgH5;0L@H81b11L0HL@1E1HAn11L0E1HE11E1HAn[HHcH5/H811E1L01HE1E1AnLYHiH= >IHpH5 H2&IHA$x A$HHdE1I9SHLH4LeHH?L@HUH)H (LH0gL@L0Ax A8MdHLL0'L0HHAALѼMcISA$=wA$=wAxAt`I1H H|eH5-L@H811L01H1E1L@HApLH0&H01L01E1H1E1E1HApLL0E1L01ApLE1 1L01E1H1ApHHH aH5,H81!1L01E1H1E1ApHL0AjE1E11E1LE1LiLL0L0KH}H(`H5_,H81荴1E11L0E1E1AiHLLL0衺L0L0AhKLL0tL0L0AbL0Aa E1E11L01E1H1LE1HA^H LL0L0pHHHѹL@L0LL0诹L0xL蛹~MZIRA=wA=wAx AfI1H@HH=bH5*LH81ֲ1E1L0HH@A^HLEHtHm^H5V*L@H81}1L01HL@E1A^H1L0E1A^HHL0rL0HJHR=w=wHxt{H11L01E1H1A^HILH@L0L0H@qHL0迷L0LL0譱L0HIAx A>IBL0LHHL0HHLL0HHxLHEL0AALHH5%L0H81褭L01L01E1H1ARHHHHHHHHL@HHA=wA=wHxt1*HJL0|HHL0$L0H1HL0L0yHk11E1HHHHH11L0APHHHHغL~LL0jL01L1E1HLE1L0AR1HHHHHHHH1H@AxAt9At:1L01E1HE1E1E1LARpL蚱H@2ISH2H=w ISHRH=/HQL02L09E11L01HE1E1APLLLLLLLLLLL0貰L0E1E1L01LE1E1AULLLLLLL頸L%A$=wA$H 1HuHH}LHE^HA$x A$tkHt'H11H0"H0xtZL0A[L0AZHL0蛯L0^LH0耯H0zHl11L0E1HL1E1HE1A^H51E1L01HE1E1A^LR/HHH5$H<HH!Hx >HHJHHH1H)HHdHx HH5 E1ɸH9rHHH4LMH?HL@HUH)H IH@H0HL0x Hx MLL0'pL0xuAx Aru/H=w HHH:H=w HHH 11L01HE1E1E1HA`HE1L1L01E1HE1A`H1L01E1HE1A`ڴL0A]fHn1H1L0E1A^HH鍴HL0,L0aHL0L01L01E1LE1E1E1HA`ݳHVHTH58H81fE1L1L01E1LE1E1E1HA`H}HmLJHRA=wA=wHxtVH1HL0L0VLlj00uHL0HH@L0辪L0H@1L01E1H1E1AMHHHHHHHH飲Hd11L01HE1E1E1HAMHHHHHHH1E1L01LE1E1E1LHAMHHHHHHHH|zHxL0*L0E1E1L01LE1E1E1LANLLLLLLLHyH5L0H81脢L0n1L01E1H1E1ANHHHHHHHH齰LHHHfInAfInfl=wA=wHx $HϺH)ELH0MLH0IApAdLLƧLH0BLL0tL0HI7Ax AIALLHH0LHILH0LLHHLLL@0L@LHILH0L0оH>8L@LL0t&Ax A}LL1LL0ANH11E1HHHHHHHH5HDž@AxAtAL06L0tB1L01E1HLANE1cLL0åL0H@L'11L0HE11E1HANE1LHLHHHHHHA1HH@LL@L0L0L@ZHHL0)ޤfoHL0LL0贤L0*11LE1HE11E1HE1A8HHHHHHHHH?HLH5EH81̝L@鬹1L01E1H1ATHHHHHHHHHLL0蠣L01L01E1LAsH鿫LeALL09AiAt LL0 LTL 1L01E1HAsL5LۢE1L01E1LArH譢]E1L01ArLE1٪H{IE1L01ArL駪LM1L0M1HE1E1E1E1Ac$I@H0Hw I@HpHw I@HPH=11E1E1H1HHHHHL0AO1L01E1H1E1AcHiL?LH0+H0,HOfUfHfHnHxHAWAVAUATSHHHdH%(H]H)`~ԒHDžpfl)EfHn)EHwLIHMgHgHt"HH=wH`HH4HUIH`L%EH(ATeY^HpH(H~-@ff.ff.HHzH<uHHLL HAH UCH5R H8S1GXZH(LxH;Htx HI9uH CvH=:E1OHEdH+%(UHeL[A\A]A^A_]DHt#HpHV=wHpH^=wHHhH=w HhHH`H(H`HpH H H HDž8HDž@HDžHHDžP=wHL(hE111HAHAHH(=wƒd+| Hm L(hE111AHHAH8IH*=wƒ6+HDž8IcE9IU HDHH H5aL*IEAuI} HHIH HLME1HHL`IMM9 JHH5LHHH8H5HI9@8HHuLHHEHEHLHESLIHx ,HDž8A$x A$,HDž@M&L5&L%M9M9;(L; .(LL_LA8Ax A-E1HCH5HHHAIMHAH5FLϺL誖LHH@ICAx A:M9M9+L;ʾ+LL荑LADAx A:E_=wH=Hu1HH]HEH@Ix CM#CH5v1LL譕LHIBAx ABHDž@M9L5ڽM9K4M9B4LL艐LJAx ABdHL9HDI=wA~fInHflHE)EAA @u tEIAH}LH[LHH@ISAxAuL誗f.HiHDž@=wHuHLuHHEIċx "HDž8Mt%11L^ A$xA$uLHDžE11E1Dž)E1E1HDžHDžHDžHDžHDžffD論8fDHH D:AHJHH<L o?H5H8S1XH`ZH(fHH9HuH;,ff.L5!HL9ILDDA=wA~tH5fInL@flHE)EA@ @u tEI@H}LHP LHI'AAL@H=wH4HuHHHEHEJIċx !Mt!11LA$x A$!Dž E1E1HDžHDžE1HDžHDžHDžHDžE1E11HDžMtAx A MtAx A MtA$x A$ HHtx Q HHtx D HHtx 7 Hd7H=fCMtAEx AET E1Htx K HtHx 9 MtAx A1 HHtx $ HHtx  MtAx A HHtx  MtAx A HHtx = Htx 7 H x / H(LxH;Ht xtHL9ufDfHpHV=wHpHV=wHhgH m5A-fDL5A=wAH`LpL H(HdsH`HhHHpH LH@  L5L9 CHHXxfDL#Mt M9H[HuE1E1L=H=IWLHH-=wHGHH5:HH.HHHx "HDž8HH9A/H HuHHHEHEHHLHH@Ax A"H#H&IH#HH HHHDž@HP Lx(MtAx A+MtA$x A$*Htx  #HHH=HDžPHSHRIH'=wA$ID$LHLH5HH(IM'A$x A$ HH=oLHSH̍LHI*=wA$ID$LHLLH5HH,LHHn,A$x A$2!HE1HI9AI-HLeLHEHELHHHI-HHHP =wHLLH)LHEHH?H HEH4HPMLH@HtA$x A$%HHDžPx ` Ax Ah HDžHAx A" H.=wHCsHDž@H{ HHHHH@i/HH@HHHH)HH$HM$H膅HHHR.fH*H=>S.HL%LLIVHH词IH.H@HHHGLLIH.HL%fLLIVHHIHH&H@HHH LHHHH&H@H;.HOLw=wA=wAx 'HLHMHHEHH@Iċx "+Ax A$HDžHM&A$x A$N'HDž@HHHHHHLE1HLH HLIHILHr@HHIf.LHMLHHIH"M9uLLHH茄IGL5LM0H=赂01LLAIHM1Ax A)Hx R'=AUxAUIE1E1HDžHDžHDžLxL5A=wALpADH@H0H HHLHHЇLH谇L蠇H萇H耇HpL`;LPELL9LDL H5aHL%H=IT$L+IHF=wAIBLLH5HH LILHM~ Ax AHAuI} H迅IHX!HH蠅LHHG$L5:H=IVLOLHHH$=wHLH5HGHH&LHHPHp&Hx H;AE1HI9D$&HLMLHEHLULHEHEzLLHIHHP =wLLLH?L)H]H HELJ4LH{LLH@IMtAx AeAx AWHx Zx dHDžPAx AaA$x A$gHDžHMoAL8=wAIELLH5*HHk'LILHMID$HXpHoH{dLHc艄LH.*L5jHLHLLFHLINjx %MLLLSLHAx A0#HlBA$x A$iH1H=HL]LHDžHH]gLHAx A#HDž8x Ax AHDž@H?&H=wƒ0wL9 HH=HSH$HHH^,=wHH5}HGHH-IL8MY,Hx $HH=6LHSH蓀LHHH,=wHLH5HGHHt.LILHMJ.Hx A%HAE1I9C/H LMLLHEHEՀLLHHS)H HP =wLLHL)H?LeH HLLJ4wLLH@IMtAx AQ+A$x A$&HDžHx (Ax A(HDž8Mo(A=wALHL1HAHH1H@L9:0HDž@ yHHXx@ff.LMt M9!H[HuHDžE1L%H=LIT$L~LHH/1=wHGLHH5<HH5HLHHHHO#x *HH9A5H HuHHLHEHEH@LLH@IAx AI*HDžHM"L~LHHH"H L` HDž@x +HDžHHtx .MtAx Aq-HHtxHh-H HH5PHGHH-IL8ME-H LߺL\yLHH@I4Ax A(L;HDž8L;àM9LL0tL^3Ax A(HDž@4HH5[HGHHB*IL@M|)HL/{LHHHI@%L{LH%L` LHLHwxLLHHHHA+Ax A&Ax Am&L%=H=HDžHIT$LFzHH*=wHCH@HH5HH`0IL8M0x c'L%H=cLIT$LyLHI-=wAI@L@LLH5HLH[.LLHHHPH-Ax A'HE1I9Cd.HLMLLHEHEyLLHH@I-H!HP =wHLLH)LHEHH?LH HH4&pLLHHHMtAx A'Hx &HDžPA$x A$&HDž@Ax A&HDž8H2.=wHCL%HDžHHHLH@HHLIVHHzHH'H@HHH%HLHH'HL%LLIVHHyHH-.H@HHH*LH@HH .HH9G2HOLw=wA=wAx .HLHMHHEKHHHIċx -Ax A%HDž@MT-A$x A$+HDžH]rHHHLHLHLE1H4HLMIHP@HHHI8LLLHH0HH I0H0H0vvAvIF D >I01:ff.H(H0HHB(HA;vrHHBHJtǀ8H(HI8A5HI(H0HDžE11E1HDžE1E1E1HDžE1HDžHDžHDžHDžDžpHtiHtHHIH2tHH@HL` HDž@8Dž-E1E1HDžtH=HXLTrLXM/nHn3DžE1E1HDžHDžE1E1E1HDž1HDžHDžHDžTLsHDžE11E1HDžE1E1E1HDžHDžHDžHDžHDžDžHs-LsCH5QH=mH1HUDHJ0H;0mHHJ0HH0VHfDžE1E1HDžAUDA$=wA$Mt$A=wALHuHPHcI IML9LLHHmHL5DH@LM.H=!$kO.1HLAIenM.Hx 'A$x A$=&= AUx BAEtHDžIE1L qyBmff.ff.HB(HH(H)03HHcHLB(L;(}ILB(HH(H0x H3XH=MHDž@% HHHH@HPH H5H9p=w H H H8IHZoHxH0pLHH^LH HLlHH/Ax A1HHDž8x HHHtx  HPHtx H@Htx HHDž@H@xH8L Htx MAA~LnqAH@x AH HnnLLZnLLFnHB0HHB(HH(H+0H0LLnL\HLmLLLmLD/LmHmHQHH)ȋVHHdHDžE1E1L8HDžE11HDžHDžHDžHDžDž(DMAALLLlLLwLLlLLlULlHLlLLLhlL}HLMlLA=AH@HlLLlLHLkLHLkLLLkLLLkL~H jH8IHsHDžE1HH@xH8L Htx =MtAx AVHPLHL@HH  L8DžZE1HDžHDžHDžHDžE1E11HDžMvAkA_LLLLRjLLL(wHHHPHڍH5LH81cLHDžM1E1Dž2E1E1HDžHDžHDžHDžHDžHDžyiH=eHXLgLXMLrcLHH&1E1HHDžE1E1E1HDž1HDžHDžHDžHDžDž0AERAELIE1E1hHDžHDžHDžE11E1E1HDžHE1E1E1Dž01HDžHDžHDžHDžHDžMAALLLgLLHfFVHH HH*HAQHH HHzLgL`gH=QHXHeLXMdeaHH%1HHL8L@HDžHDžE1E1E1HDž1HDžHDžHDžDž\HDžE11E1HDžE1E1HDžHDžHDžHDžDžSH H1HHHLH@H; E1 `LICHDž_LINLej_I_LHEE1HDžE1E1E1HDž1HDžHDžHDžDžLLHeL|HL-eLLeLLLdLLrHDžE1HDžHDžHDžE1E11HDžHDžHDžDž14HxdxLkd)HLH8\_Ax AL8L@HDžLHDžcHDžE1E1E1HDžHDžHDžHDžLcfLdH=HXHaLXLM+L]LHHp&1E1L8L@HHLHDžE1E1E1HDž1HDžHDžHDžDž\`LbHbHcH=HXL`HXH\HHLH5JH81&\MpM`AfIn=wAA$=wA$Ax AHuL)EHHEWIAALLaLHDžE1HDž[HHBL8L@MHa{[LHLDžE1E1E1HDžHDžE1E1E1HDž1HDžHDžHDžtLGaH=اHXL_HXLHHZLHH1E1Hf.MaIYA$LP=wA$=wAx A I1kHqLyfHnH =wA=wAx R HuL)E5HHH@H_L_LHLL@HDž\L8E1E1HDžHO_LL;_LbL'_+HDžE1E1E1HDžHDžHDžHDžDž1HXLHL^L>mL^LH^HML$Mt$A=wAA=wAA$LHx A$ME1DžZL8E1HDžHDžH@IfHDžE11E1HDžE1E1E1HDžHDžHDžHDžDž2-HLn[LHDžE1E1HDžHDžHDžHDžE1E11HDžHDžDžUxuLZLH@HDž8Htx Hm;H=LHDž@X HH@H8HHH H~LHL@H9BL8Ly=w0HxH HH MtA$x A$ HDžHMtAx A HDž@MtAx A HH@xH8LHtx lHtx `HHOE:H0Y-1DFL%Ax AL8E1E1HDžLLXL HDžE11E1HDžE1E1E1HDžHDžHDžHDžDž fRI@1LLWHAx A>H2HDžL8HDžE1E1E1DžcE1E1HDžHDžHDžHDžE1E1HDžHDžqHLWLL~WIE1E1HDžHDžHDžeQHHAALWE1E1E1Dž6E1E1HDžHDžHDžHDžHDžE11E1HDžE1E1E1HDžHDžHDžHDžDž L8HPE1E11HDžHHDžHDžHDžDžELLULHDžE11E1HDžE1E1HDžHDžHDžHDžDž!CLUHLwULME1UH=?HXHxSHXHH~LOHHDžE11E1HDžE1E1E1HDžHDžHDžDž6|HDžE1Dž LTLLTLXLTLLkT-LTH=;HXHtRHXLHH ANLH1E1H11E1E1H1E1E1H1HHDž6LLSLHMI LLSLLLmSLLYS8HLASHLHLSLL)SfoLRLHLLLRLLLH)Rfo1HyHrLLI}L[RL8LHH=[QLLRL\LRmLLQL HLQLHHQHLCHKIMKI[A=wA=wAH8x A IE1H=AEAELIE1Q1H鏽LPL LP-QH=HXLNHXHJHH111L8H@E1E1E1HPHE1HHHDžF@HuLH8NK1E1E1H11E1E1DžKHHH鱻HPE11E1LE1L8LE1H1HDžE)LH OH tHH OH }HP11E1HL@E11H1LHHHHDž@ݺHHL OHL HHIH)HHHRH9HXHH~H~1H;D |HH9uHHHRH5HOrH81-HE11E1E1LE1E1LLE1HH1LDž8L׹HE1E11۾H=d1E1L@H1E1LE1LLE1HDž8n1E1E1E1HDž!SLLLMLLFLMH=IHXLKHXLHVGLHLHrLH5H81FL%=wH@{LLLLLLLLLLLLLLLLvLLLLH=?HXLxJLXLMLEFLLHH HPE1L@HE11E1E1LLE1H1DžFELLHH1H1E1HE1E1E1DžF1HHMKI[A=wA=wAH8x An I1RL KLJL;HLLLJHLLLH LJHPIE1E1DžFE11H1HH1HRDIE1E1H5xH11E1LHE1E1Dž4HHLoHZE1E1E1 H=1111H1E1E1HHHHDž42GHuI HnLH8nDH7HLHL@E1LHIHHH LHBLHHxHPLHDž@H1E1E11H1LLE1H2HPE1E11Dž@LHH1HH1H1H$LILAAfIn =wAA=wAx HLǺLLL)ELLH@ILAALL}GLLL%3A$=wA$H HkLHHBH9| =wH HH5lH`IHH@H9<A=wALPL˅x A~HHfHn~›~ fHnHEflfl)M)EF @mEC @ZE HHHPHCHTDHHHx x L1HuLHPHUHLuH@HAx AJ A$x A$ 1HHHt&11H|x 1H@1E1E1DžAHbL EEHDHLLL)DfoLLLL@fHuH?)EL¯HHIHLE1QD1LE1H1HH1DžKHƯL@LH11E1E1H1E1E1HE11HHHHDž"$HuA]H;[b{H5LL;LHmHP1E1E1HE1E1E1H1LLDžCcH;bH PXHHtHP1E1E1HE11H1LE1HDžBH;=pHHH9HuH;aoH`H5H86>tHJbHE11H5LE1H81w6E1E1HLHL@HL81LE1LLLE1Dž\HaHH5$H816(LCMt I9 MvMuHDž(E1H IH={HQHH`w3IH=wAEIEH5LHH7IMAEx AE HXI9@nH@HuLHHEHEL`ٞH`HEIŋx R MHP2HH3HHH`Lh HP(HEMtAxAuL2MtAxAu L2@H(Htx  L-ԊH=yHEIUL1IH=wAIGH5LHHIMAx A7L-XH= yL@IULf1L@HI=wAIGL@LH5oHH<L@IMAx A HEH`VI9BH`HuLLmHHEHEL@誜L@HEIAEx AE_ Ax A< MA=wAIFAvHEI~ H HqwH(HP> HX=wLX1HH=HHEL}k0HEIAxAu HXi0MH51L,IH@AExAEuLL@#0L@L;THEL;cTXL;0KLL@'L@AAx AE L-χH=vIUL.IH=wAICL@LH5DHH/L@HHAx A9HSH9A`HXHuHHHEHEH@;L@IAx AML;0HPIIfHH0ff.ff.AXHsff/:HxH5LHLHIUHH@$0LHI H@H@HHH-LLLH@H@LHIj"HxH5L@HLHIUHH/L@HHAHpHHHHLLH@H@LHHEHHpH;5Q LWLoA=wAAE=wAEx HLLULHL@HEL@HLAx A`AUx AU9HEHGx XLH@'H(H@HLAH0 HPH L=HxLHLIIHLn1HL@HLE1I@HIL90BLCLx+xff(YC @fB.d\YzuH HHLHPILTH9(ELLLHHLx%IGL-+sLxLMOH=Lx$Lx1LLAI?'MLxAx ANAEx AEA=HXAxHXAMfDAMcAh AH`tH`xH`OAx AMLpI<$Htx IL9ufDHMHLL FAH H5ǔH8AT1"AZA[ f("fDHNH0=wH0H]HEH@HEHp@fDA=wAMwA=wAL+H((HP'HEIH(H`HEHH`HELx kfDH(LL`'L`#DL'L'LL@'L@H'DA @'*fDMAA H`H`H&'LLx'LxMq A6=L&kLL0LPH@Lx&LxH@LPL0SL0LPH@Lxw&L0LPH@Lx(fDLL0LPH@Lx,&L0LPH@LxLLPH@Lx%LPH@LxHL@Lx%L@LxLLx%LxDHx%Lx@H`HH`H`0%HX%AYMAVL$?AE11E1E1E1E1HDž`HDžXvfDL$E11E1E1E1E1AHDž`9H`HHS$fD$H=$kHUL`"LeM}>HDAE11E1E1HDž`E1E1fH HHLHPILTH9(IHxL-vLLIWHH@e%HHH@H@HH* HLHHHxL-vHELLIWHH@$HH^H@H@HHN LHEHH@H@H;lGLHOA=wA=wx  HϺHL}H@HE蔍H@IAx Ax  HEM AEx AEb HH(0HPHL HxLH5HLILnI 1M@@ff.HPfE1H0H @BLxxCIXM9u@HPLH0H ^Dff.ff.HY@H9uLIHH9(DLHHHL- iH@LMMH=d 1HLAI-M$ Hx  A$x A$* AE1=HX HXHLxMLxDD H=fHULHMHxHxHxHE1E1E1E1L`E1AfDH`D;H`I:@E1E1E1AHDž`fLL@L@DH`H[QH`?M~MfAfInX=wAA$=wA$Ax A HL)EH|HE萉HEIAALL`L`E11E1E1HDž`ALH`HKAH`/H`H`H`#@M}fInMeAfInfl=wAA$=wA$AEx AE; HL)EL`PL`HEIAALL`L}E11E1E1E1AHDž`8H` H`#@HXvf.L%cA$=wA$H@lHuLHHEHE^IA$x A$M11LAEx AEAH}HEHtx E1HEMtAEx AEH5H=2H HuHMHUH@H?H9F=wL@HPIHuH`TL`HITLP HLH@L@HH`AEx AEAx AH}Htx eH}Htx 7H}Htx 'H HEH@xH8L8Htx M6A+AL-H=`HULL}MGE1IH*1E1E1E1E1AHE1H H@xH8L8Htx  MtAx A H(Htx  LMMHDž`E1AAxAtL}E11LL@LxL}E11L@Lx^1E1E1E1E1AEIEH`H=m_HULmMdH 1HEfDL@H=%_HULaL}L@M)Lx1LxHXHZ=LH5H81Lx1f.L HEI1E1E1E1A3L@IH<LH58H81E11E1E1E1AMPIHAfIn@=wA=wAx AHϺH)ELH`4LH`HEIAIA=LH`)IJMzfHn`=wA=wAAx AHLLmH@)E萁H@HEIƋHH=HXLxMALxMAMLL@LL{wH}#t(H}HEHt HEHHLL@L@%LH@H@E11E1E1AzL)`fo`GH@H@'LH9LH8HLMxHLMAMkE1E1E1&1E1E1E1ALLL`)0fo0L`A1E1E1E1XH8LH8AE11E1E1E1-LmH@:H@IH,E1E1L6;H%; HuH6H5H8@HaVHILE15A=AH=YHULL]M Lx LxHE11E1E1A f=wH}!E1E1E1A L@HMHMfHuH)EHH@0}H@IE11E1E1AsLiLyAEfInX=wAEA=wAx $HL)E|IAEpAEcLL@5L@HLLHH@ LHH@LLH@LLH@HWHIyH9HXHcHqH~1H;T HH9uH34HJH5LxHWH81 LxMA1E1E1E1L1LLWIAxHLM_DHLH@LH@HPH(1fff.HHH9LH()@zH(fo@LHL`)IfoHL`(Hg4HLxH8! LxAx AH}AHAE11E1E1L`LxL`Lx8LL`LxL`Lx'LLxfLx3HL`LxDLxL`H}HEHLH@LHH@wLLx LxE11E1E1HDž`E1AgAx A]AE11E1E1>H2HLxH8yLxHLxQ Lx1LLLxq LxIAx AcME11E1ALmfHuHLLH@)ExH@L6LxHH!0H5"{H8 ALxAL\ LxH)@A fo@H1LH5E1E1AH81E1E11L`E1E1 H H0H(HPDH9HDH9H H0LH5WHxH81,HxLLxa Lx^E11E1E1ALx HHH9cHuH;/QH.H5LxH8' LxH40H`H5H81l1H}LLx Lx1L}E11H`AH/LH52LxH81Lxff.UfHZfHnHxHAWAVAUATSHH`dL$%(LeI)E~HEfl)EfHn)EH=LIHM-MtIH=wHUHHMHUML5&HhJ4AVxA]A_tdH}}H}JL]M%MH,HLL BAH H5tH8j1[A\HhLeH;Htx HI9uHH= E1茸HEdH+%(0HeL[A\A]A^A_]I LA=wAL--L]AELm=w P@EAEHELmML0HhA=wAA$=wA$L-2`H=NLXIUL@LXHI=wAI@L@LLXH5XHHRLXL@HHAx A# H4,H9C HuHL]IHLXHErLXIAEx AEMAx A L- ,M9 H,NI9D$yHZMH5wLID$H5 bLHHJHHLIFH5aLHH=IM<LHLXLXHH9x Ax ARH;*H;a* L9 HHXHXA x  LL=]H=LIWLIH=wAICLXLH5ZWHH(LXHHAx AH)E1H9CyHXHEL}LeLuTHHH)VH=_HXHP =wHH}HH)H?H@H HEH4HXMH@HtAx Ax x H:.x L90B H;\H=JHSHXIH=wAH\H=JL@HSHL@HH=wHALPHH@H5'[HHH@LPHHx JH(H9C7HuHLeHL@HEhnL@HINjx MH'I9CH0HuLL}HHEHEL@mH@IAx Ax rMALL@L@Ax AkxID$LH5WHHIMI@I@HH HH)HHH0LL@&L@H0H0Ax AID$H5\LHHIM"Hcx11LL@yL@HIAx AjIC ICHHHH)HHHLL@2L@HPHPvAx AID$H5\LHH7IMHcx11LL@xL@HI Ax AI@ I@HHHH)HHdHxLL >L H@H@Ax AH FHxLXHHL9x H0IHID$L LH5QHHZ!L IM6!I@LLL H5TSHHy!L LIMQ!Ax AL;=>#L;=#aM9XLL L A7!Ax AH`H5QLLHIWH HE-HHI#H@H LHHL LLL HIDH`H5PLLHIWHH HH#H@H LLML LAHHH@L H;"8#LGLA=wAA=wAx @ HXLLLEL HE*hL LAx A!Ax AHx 8 HCH9C iH`LXMIL`L0L HXLP)I$(ID$(I$0I9L$ HLL1M$0AWHL@*CID$At$HHIL$AYAZtA$8 I$(=ABHR8HR(I$0|fDHaHH1L AH H5#gH8AT1XHEZHhQDLLXALXHH9CoHELeH0@L- AE=wAELmfDL-AE=wAELmafDL{fInLkAfInfl=wAAE=wAEx,u%HLX)@]fo@LXHuL)ELXeLXIAAL LXDfDA=wAfLuE"HHHRH5GHP =wHEH=sKHuHHHHXKIAx Ax tMQAEx AEMfDCLLXLXDLIFH5mPLHHKHH.H5%WHPIH9xuHL@xL@L;L;M9LL@-L@xAx A2M9=AD$@H}> 1LhIH L9H`=wA=wAfL}EIH)HPH EHP =wHXHMHLH=Q%HAx AAEx AEH`HfHXHM1HH=TRH]H`Iŋx Qx 11۾MAEx AEH=Lp D!=Et A$=wA$MLXH=%=HxL^LxLXMIL`+L`Hff.M޾l1E11/1E1M޾lMtAxA3Htx HUH=tWHtx MtE1AxAtzMAxAtxA$x A$tGHhLeH;Ht xtHL9uMffLfDLyL{HBHω``IE11۾vf.LljXH`kH`XfKL@LXHH-APH)HH0dHL@L@sLL@L@GLLL@L@YLHPHPH@!tx  1E1@4AxAt1H`xH`tN1f.LH`xH`t1ɾ1?H`X1X1fLHXHXDHL@HXL@HXRHhL-!9AE=wAEHhKHuLHHEHEn]IAEx AEMt11L%dAx Ar1E11!HL@L@qDLqs1E11LL@yL@GDHH@YH@DHH@9H@DHL@L@DHL@L@Mo1E1E11MtAxAtOMAEAELXH`XH`fDL߉XH`cXH`@KH1E1t,I1Ҿtx uE1HMcAHE1RE1E11H`=wL`HX1LeH=KHL}|HAx AH1E1Hx  A$=wA$MD3 It$0I$0I;$0=HI$0It$0I$0tASH)HHP#LCL-5AE=wAEHGHuLHHEHEYIAEx AECMt11L`Ax Au1E11APH)HH@_A@APHH HH0YLH=4HxLLxMWHHX1E11۾v@LL@L@{A@APHH H0LrH=I4HxHLxMU]Hk1E11۾sHHIH@H LHHL LLL HIzH`H5?LLHIWHH HH H@H LLML LAHH H@L H;LGLA=wAA=wAx HXLLLEL HEWL LAx AAx A}H, x LXALXHHCH9C NH`LXMIL`L0L HXLPH0%@I$(ID$(I$0I9L$ HLLHM$0AW1L@2ID$At$HHIL$_AXtA$8t'I$(=1HR8^HR(I$0넃It$0I$0I;$0 HI$0It$0I$0B1E1E1vW-LXHL@bH=0HxH,HxL@H:LXH`H`LXHE1E11۾fL{HSA=wA=wx 4 H1FLL@PL@ACASHH HPACASHH HHPWMAx Ax G1E1v5L#E1E1LPH@H+HL@E1E1`HXuHX`1L@H`=wA$=wA$fLeE}HH HAH5s2HC wHuHHXHH=QBIA$x A$ x  M L`HX1LmH=BHL}LHAx A AEx AEi 1E1HyfLHcR WLCfInHKAfInfl=wA=wx  HXHϺL LPH@)EQLPH@IL AjA^LLPZLPH@<1E11۾[HcR LL`L MLXH0L`/IEH5t.L`LM H=NTLXH`CY1H`LAIMLXv AEx AEAx ALA$=wA$H5LL`L MLXIEH5-LM H=kSLXH``u51H`LAIMLX!HAEx AEQ 1E1A@APHH HH@A@APHH H@MSI[AfIn0=wA=wAx AHXHL@L})E2OL@IA+ALL@L@Hcօydff.ff.ff.ff.ff.ID(I+(HI$09I|(I$0I;(}HcHI(I|(I$0HcօyAff.ff.ID(I+(HI$0I|(I$0I;(}HcHI(I|(I$0YL@L@H1E111E11۾@I!1E11۾wLLL L :LL@`L@H05L@L@H'1E11I@1E11۾LL L HoA$= LSl1E11۾8IL@DL@Hn1E1E11%LL@aL@HP*I$(ID$(I+$0ID$0I$0I$(ID$(I+$0ID$0I$01E1yrHH`E1xHھyLH`E1AE= AE1E1+L ILnLL`L` C91ɾvLL I HھLE1HHX&HXLL HXL HXZLL HXL HX"LL LL HHLXLXLL gLL [HLXELXHLHPL@) Lfo HPL@.ML`11E1LH #H I|$H9HXHHqH~1H;T HH9uH?HJH5LHWH81z1E11DL@!LL@)0%fo0L@ H MLL HXL HX1LL HXL HXLLL\HH8*HoH`^HH8ZfHuHL HX)ElELXL _fHuHL HX)E0ELXL 11۾~HHH8YHLLzGLm^HLH5!{H81L`QL91LjL`IAEx AEM1ɾFfDH1 HHPH9HXHMHqH~1H;\ HH9uHKHaH5I1HRH812M~1E1Lh1LL`L`IAEx AE M1ɾn9HHH5FH8:AEHrH5zH8BIH"HH5yH81^uHHH5byLXH`H810H`LXHLH5&yE1H811ɾvLL`4L`LL`L`HHH9HuH;u7?HHH9HuH;1H8H5QyH86H=-111H1HHH9HuH;_x@]H3PHH5xH8XM1E1~;HyvH=;UE1軅HoH5pDH8AE%-@ff.UfHH&IfHnHATSH@dL%(LUI)E~HEfl)EHLIHMIIM$HL]H]LeHzMHLPJ4LU>D_AX6H}LUL]M"J|IIt J|HuHUVfDIt IHV=wHUH6=wHUHuH]LeHL HH;Htx HL9uHEdH+%(HeH[A\]@MH wAL |HHH]HyH5~?LeH8AR1nXZH;Htx HL9uHUwH=RR1PHV=wHUH=wHUH vAL G<H1=wHU#fHHH vH5>L ~AHxH8AR1wY^fDHMHMjfDH1=w HuHUH]LefUfH*fHnHAWAVAUATSHH8dH%(H]H)EHHP)EfHn~/HHflHE)EfHnfl)EHt.LIHM~"H HHcH>Hl HL Hz HV=wL&HUA$=wA$LeH}H]HEL}HHHHDž`HEHDžhH@HEHDžpHA$=wA$=wH5"HY9 x HDžXH Lpx@ff.M.Mt L;-3 MvMuHDž0E1H U)H=&HQHH(HHg=wHGH`H(H5$HH=H(HHpHzx  HDž`HyH9AH@HuHHHEHEH(:L(Ax A HDžpHx  Mt#AxAuLff.MtAExAEu LfH0Ht$xuH@ff.H5)-H@1[HhIHH;L;5 L;5 LWAŅAx AiEHHhIHoH@HIHgAx AL;-dHDžhL;-/ L;- LAƅyAEx AEHDžXEdH@H@H@H@HHYHH)HHeHeH@AQHH HH(LhxfMuMt L;5MmMuHDž E1H5$H=HVH@HH=wHAH@HH5HHH@ILhMx HI9BHuLHHELeL@O7L@H0Ax ASHDžhH03A$x A$MH0H5:#HGHH&HHh&H5)H9HAH;U5'HQuHHx &AH Htx }&MtAx AK&MtAEx AE>&EL%"H=HDžhIT$LIH(=wAIBL@LH5HH(L@ILXM(Ax A_%L%e"H=IT$LyIH(=wAIBL@LH5HH0)L@IMG*Ax AY&HE1AI9Ez*H0LEHEL@HEHH*H$L@HP =wLK4LL)H?LuH L(H@3L(H@H`IMtAx A&HDžhAx Az%x %AEx AE|%HDžXM&H0x %ID$H5#LHHG*HH*HH0GH0HH@+x %L/HDž`M9+ID$H0H@y.L0L@11Hff.HHI9]HIH)H9~IH-LYA=wAHL@L@HIQ1H=#HL8H@L@L8HHhI1Ax AT+LHuLmHL@HE&2L@IAEx AE'+HDžhAx A2+M0L11L@8L@Ax A+Dž@ E1E1E1HV=wHUHV=wHUHH iAHHHkL nH51L}H8S1zXHEZHHHHH;Htx uHI9uHOijH=)EE1IuHEdH+%(5HeL[A\A]A^A_]@HV=wHUHV=wHUHV=wHUH=wHUHHML}IL%jHHH4LAT4A]A^H}H}HHH~,ff.ff.HHbH<uH,HLL UmAH gH5/H8S1[A\r#fDH5iH,AŅxuHHDžXE1H =wH|HuHHHEHE.IƋx M!11L5Ax AdE1E1E1E1Dž@AE=wAEMuA=wAL:H0fDH fAfDD{LH(H(XDH(H(HKH=w LeHEH]L}HHLYA=wAL]DsLeH]HQ=wHU7fLA=wAL] H=H(HxѿHxH(HqH@螻HuHH@H5.cH81 HXHtx HpH`HDžXHtx HDžpHtx H H5HDž`HDžhH{p IHsdH=M@ppHHhHpHX#DH=L=wHHuHH@HEHE+H@H`HËx uHt,11H=2xuHHDž`Dž@HXH H@xH8L(Htx EMtAx AeH0Htx ;E1H7E1E1E1ҋx JH`Htx ]HhHtx pHpHtx MtAx AMtAx A@HbH=g>nMtAx AMt1AEx AEtZIA$x A$t6HHH;Ht xtHI9u@CL8fDL(fDLqHL0L8H`L0L8HL0L8ʽL8L0zL0L8蝽L0L8jf.L0L8mL0L8Wf.LL8AL8JDL(PE1E1E1E1Dž@DHHDžXILؼDž@1H@#HH蹶HE1E1E1E1Dž@8DkaHTbJ7H=L0Dž@E11E1LE1IA=wAIvH =wLI H@H=HxHxH軵H+L0E11Dž@HDžXHDžpHtx  HDž`MtAx A H(H5HDžhH{pkC @H^H=:jHH`HXHhm>HhHtx HXHtx H`Htx tH(HDž`H@xH8L0Htx LH Htx Mt"AExAEuL fH=wHHhEIHhH= HH@胰L@HIAx AHuHLmHHE$IAEx AEx jHDžhMt!11LV+A$x A$L0E1E1E1Dž@E1H@H)ȋVHH!LŸL踸$讲H(HHDžXLYLAAfIn@=wAA=wAx LL)ELL(x#LL(AALHHL(qDž@E1E1E1H@FVHH HHHL@藷L@RL胷E1E1E1E1Dž@&L[H=wHHuHHHEHEh"HhIƋx SM11L)Ax AZE1E1E1E1HDžhDž@rDž@E1E1E1ZDž@E1E1E1Bx H0H5HGHHdHHH5HH@ӲH@HHhHdx {H;qH;GH;HH@谭H@A&x wHDžhEH H=[HQHH@踴IH=wA$ID$LpLH5 HH8IMA$x A$H0L@H5HGHHL@IM5HI9BH HuLLeHHEHEL@L@HhHA$x A$HDžpAx AHH; H; dH; (WHH@H@Ax HDžhEH H=HQHH@H@HIs=wA$ID$LpLH5HH&HHA$x A$cH5H01H@˯H@HHpIHbH@HH`HH0HpH@HHHx !HLHH HHHH@H`0A$x A$CHDžpx BHH9AH@HuHHHEHEH5LHhHH@x HDž`A$x A$MHH; AH;Du+H;t"HH@QH@AċEH-DiH=wHHp\HHkH=) HH@蚧H@HIBx RHuHLmHHEIAEx AE%x !HDžpMt11Lo"Ax AE1E1E1E1Dž@H H(H@xH8L0Htx  H Htx  MtAEx AE L0E1E1E1E1QHL8L8#Lk9H@vHIH@IL0Dž@IJfInMBfHnfl=wA=wAAx ALǺL)EHL@ HL@H0H艮L@y-fDHXH(HH5HL L 6tH(ɚ;H5HL zL tHH9(HH9(L9L-H=sL IULЬL HIF=wAIBLLL H5HHL LILhMUAx A*HI9F}HHuLMHL HEHEL IAEx AEHDžhML;AL;DM9LLL 8L LAKAx AElHH H9H&=w HHHhH@LIH'H L HI&L@ HH蒨IH/!HLx Ax AMH5bH=LHVH lLHI=wAEIELhLL H5 HHL IMAEx AEH5H=uLLHVH ǩLLHI=wAEIELhLLH5HL HL LIMAEx AEHE1AI9B*LLLHELML LuLLHHhHH)L LHP =wLK4LL)H?LEH LLLHL LL HLIMLtAx AHDž`Ax Ax HDžhAx AMTH@IEL LH5HH=L IMH@L L6L HILHLH躢LLHH Hh\ALx AAx A>H LlLHH`!H xH HDžhM9!IEH5'HHH8H55He:H H!HH59HH HH9C!HKHS=w=wx  H1HHHMHH?HH1H I4HMHHhHQHx NH4 x H1HhH8HPHyH0LLH(H@HXZРH5H 1+H I‹x MdAx AWAE=wAEAx AvLfDDž@E1H0HiLLݤHФLä1L趤^L詤L蜤H菤vL肤lHH@nH@jHLL()EfoLL(H;[HϺH@蹠H@HHwH;oH;=EH;=HH@誛H@HAċx ExuHzE5@L`LH@LH@kH8oL+wH|HoHܢLϢ$ŢH踢0DH袢H蕢rLDž@E1LsHfH`E11E1Dž@ML0IfDL(H@)gH=HxL1LxMY HDž@1E1H`IE1E1L0e諛L@I5Dž@H`E119Dž@E1E1E1,Dž@H`E11虡H=*HxLcLxML@7L@HyDž@L01E1@H0ܚHLH@ŠH@贚L@IDž@x HpH`LhHDžXHtx HDžpHt xtiHHL8 L8HH8HpH`LhH8YHL8迟L8|LH8蠟LH89Dž@L01\L0E1E11Dž@E1 HLL?H2MEMeALh=wAA$=wA$AELXx AE ME1.D8-H@谞H@Dž@L0LM Dž@I1HHH@^HH@LCL0E1E11Dž@E1HDž@1dLHHHH՝HHH躝HLHL@蘝HL@+LH@vH@.LL@[L@LLH@@H@LL0H@)foL0H@H=H It$H9HXH]HyH~1H;T HH9uHpHJHVH5 H81?E1E1E1E1Dž@ 7@Dž@E1E1E1H0lH0H E1L`L9M9ID$H0H@LHH4 H 7H LHI HP sE|LL@蜛L@LL8L@zL8L@LL@XL@HDg1UJf.ADAH@aH@H=Hx'LxMgHDHDžXLhHDžpDž@]L襚=Dž@LhHDžXA$x8A$LL8[L8HDžpHDžp/HDž@}H0 L@I/Dž@_IE1E1E1E1Dž@ H@H=HxחLxMK貓HDž@E11DH5HL kHDž(L 1LQLaAfIn@=wAA$=wA$x ^LL)EL$LHhHAALH蝘HHZ=wHE1HHuHLEHEIƋx [M11L[ Ax A@E1E1E1E1Dž@E1E1E1E1Dž@E1E1E1E1Dž@Dž@Lh13Dž@Lh1MZfInMBAfInfl=wAA=wAAx ALǺL)EHLL@HEqLL@HhHAALHHL@aLLL 躖LL H@L贎IH7H זL HHhIHfInLfHnfl=wAF Dž@Lh1Dž@E1Dž@E1E1Dž@ E1E1E1LL ÕL LLL 衕LL LL L WDž@LhLL@NL@b=LL IL kH=HxL5LxL ML@L@Hk Dž@$E1E1E1}Dž@$E1E1hLnA=wAHL@L@HI H=H&L@HIP AEx AE E1LHuLuHL@LMVL@IAx AAx AM11LAEx AEE1E1E1E1Dž@THL){foLxHL XL LLL@)/foLL@LXH H=HxLxLMH1E1E1Dž@(HhnLL 蜒L LH@脒H@HA=wHE1HHuHLUHEIƋx  M 11LBAx A E1E1E1E1Dž@Dž@$E1E1E1Dž@$E1E1E1wL0E1E11Dž@#E1K$H HP H8H5pH$H`HH HH9C HKHS=w=wx K H1HHE1HH?HHMH I4HLU'HHhH}Hx  E1L`H x  1HhCH8VHH0HLH@H(L6_HAX[H5H 1H HËx  HHՏHYH5r2H8&0L訏}HH@H5X1H814Dž@1INMnfHn=wAE=wAEAx A*LL)ELH UH LI<1HL ЎLL Dž@(LE1E1:MJMjAL`=wAAE=wAEAx AME1Dž@(E1?LL I"LHLLL HLLL LHLL ǍHLL HLL 藍LL LL uL Dž@.LLKLHL0LL0LE1Dž@(4LLLLL4Dž@(E1E1ņL ILLH H=}Hx蹊LxLLML@L@HDž@(1E1HhLLLL LLL LLH )LfoH LLLL 訋L LL-HLH5G-H81#L@`HLH5 -L@H81L@gDž@E1E1E1LhL L@LDž@1Dž@E1E1H&H@H5,H81^ L衊L@gHH@H5J,H81&HHH9[HuH;ZILE1E1E1Dž@'HH H5+H81轃?L ILuA=wAH(L@:L@HIH=HL8H@jL@L8HH`H3Ax A1LHuHMHH]H8L@H8L@HhHËx Ax AHt'11Hx E1LhDž@7E1E1~Dž@9oLH8L@薈H8L@Dž@.E14HʭH H5&*H81L@Dž@9LMH)L@LH)Dž@2E1L0H݇LЇHÇHH诇HL@蛇L@L臇.1ɸpH H H`tLLL0ME1Dž@;E1Dž@;E1E1Dž@1E1E1HH8LE1ML0H8E1Dž@;H`pHHH誆HHH菆HH{HnnDž@E1E1E1Dž@%E1E1E1H5zLDž@/E1E1L!LHDž@.E1E1H xtDž@2E1E1vH 觅1ɸ HHH胅HHDž@%E1E1HLH5'H81~XH3L&"Dž@E1E1E1HL &L HHHhE1E1E1E1Dž@'Dž@'E1E1kDž@'E1E1E1Sf.fDUHATISHHHHtHUHAHUu H1HtHLH[A\]fH[A\]HHHtH1DUHAUIATISHHHHt HӅuDI} Ht LӅu2I}(Ht LӅu I}H1HtHLH[A\A]]H[A\A]]DUHAUIATISHHau%HpHtHLL[A\A]]H[A\A]]fff.UHHHH"HHBwHt xtIHHwHt xt 1fD裂1HU菂HUHfUHHHHt,=wHzxt=Hr1DH Q=wHzxt*HJ1fHuHUHuHUfHUHUH f.HݦHHHP=wHt xt 1UH臁1]UHHHHHHBwHtx Hz HB wHtx Hz(HB(wHt xtwHzHHtHBHxt 1D軀1HU觀HUHZHU臀HUHgHUgHUwff.UHuO10Ht-H CHHwrHD։]H]HH518UHu/10HtH äHHw]fDHH5z18HGw HG@HGw HG@UHHHGHt Hx H=HE.HEff.UHHHGP8HtfDH( 2H=HEt.HEff.UHHHcd}Ht@HLH=HE$.HEff.UHATSHHHGIHUH5HHtzHUHHtGLHQzx+xt1H[A\]DH}fDxt5HH=m-fDwHUHfHx}HGH5HPHH81w@UfHfHnHAWAVAUATSHH}dL$%(LeI)E~@oHEfl)EHLIHMIIMHH]LmML5"J4HLAV_AX H}<M"J| IIt J| L}LuUIt IpHV=wHUL>A=wALuL}H]LmM A=wAIGHHpuHH-H9t]ADž`x1#WLEILHpLU]LUHp Dž`yHuLL`BL`IDž`zLuE1Dž`zLu1E1E1Dž`q~VLpI\H=KHULZLMMeVHDž`q1E1HL`Lp)P\foPL`LpL[E11Dž`vIJfInMzfHnfl=wA=wAAx ALL)EHpHpIƋ-"HC[L`MDž`vLLpLM[LM1E1LpMAAcDž`n1E1LLPZLPcLlj`Z`LLP`ZLP`NHHUH5+H81TLu1E1Dž`k/HUHEH5HH81SLuLE1Dž`eH_LH5H81SLu1E1E1Dž`iLLLMH`Lp#H`LpLMIkXLHp)`~Yfo`HpH~LH5#H81RDž`q1E1+L`LuE1Dž`vIHw~LH5H81RE1NHPHH~LH5H81RLuLM1LpLPDž`vLuIE1E1E1nHpH}LH5JH81&RLEHpDž`vLff.UHSHXdH%(HEHH\HEfHn)EHLIHMHt HHwHMHHMH4IHHUSY^H]HuH|=wH]H5ɨH!THCH5HHHJHH,H5H}XH}‹xu UWUMUH==w H=ўHuH]HH}HEH}HËx HUH}Htx VHEdH+%(HH]f.HHuH=@HH E1L HyHH5H:PH1OXZH}HtxuUfff.H}H=1+HHCH9t0HXHtTHqH~k1HH9tWH;T u=ff.ff.HH9tHuH;0ytfDH AH=HQHHMRTHHp=wHBHUHH5.HH HUHHbx #HAHMHH5IHHHMH‹Hx HHHURHUx H5HH=HVHu\SHH=wHAHMHH58HH%HMIM xuHLESLEA=wAHCLEHH5HHLEIMbH=N1HuHLELELMLMSLELMHAAAx AAx AH=3ƒ8 HRH ^AL yRRHuH=^HUPHUHx{LHfHH=%1+RfDH =w H H5H=֘HMHVHu5QHMHH=wHvH9GhHuHMHH]HEH}߼H]HMIxuHLEHMgQLEHMMHHuLEHLEHMHE|LEHMHAxAu L QHMxuHPHCfDHH=JH]JLEI1fHHMtPHMHHU\PHUH5PH׉M%PMD;fDJHAx/AuLLEOLEAxAuLOpIHUHRGHO:IHMHAfLLE]OLEXLLO^LLMLE7OLMLEuOHuH=HUAMHMHIH1HOtHUH5H81HDHMOHuH=HULH}HuHMHHMHuHHMH3H+!HωuVNuLOfHnH_AfInfl=wA=wxtsHuкHLMHM)E@LMHMIAVAJLLEMLEHM15GHMILMHM)EMfoELMHMjKHrHUH5*H81GHrHUH5 H81FHMff.UHSHHHGH~HDH{HtHCxtHHHtHǃxtHCHH]H@LݐLfkGuHSHHH9B0`HEPH]UHSHHHGHuRHCH{HtHCxtHCHH]H@KfFuHSH|H9B0uHNEtH]@UHAWAVAUATSHHdL4%(LuI 10IMHL^p~~=fHnHxH-x~ l=fHnHMUPflfHnIEAfl=PMUXDAfHELc)`)p)U)M)EMt0LFLoH~I:H wJcH>II|I HS0=wHpHS(=wH[ Hh=wH`H`HXHx HhHnH9G#HGHƒHGAI)LIUHxH}LpHPHsH;= nH;=mu L9‰HHKH5mH9L9ˆ@ M9A=wAL92HCH:AE0MMH4MkIVA=wAH=Û1HuHL8fIn()EHL8LmIAx AMAx AIT$MA=wAH;lM9 /aI}Xx MuXM9IF Ic}0L8IE(H~EIcU0L8HIE8HIU@.=wH{ L8MI1L $DMIF8LH`ZYHXHXLuH;Htx HL9uH:DH=AEx AE E1HEdH+%(HeL[A\A]A^A_]L@\L@LTiHA)HQ=wHxL8CLhL8=HPHx uCfff.HH=AWAx AHXLuH;Ht xtHL9uA@3Cݐ+CKfDDžHLEHHt8H0t?H0INjBHLML8HPtA$E1<A$MAALP+?PHQ=w H`H`HxHXf.HH)HHnH L4;I@HbH518I-f.H %A-fDH`a@LP>A$A$LL ML8xuHL8=L8H5HPL8ZL8%H=wI}Px HIEPAE0~LHcIu@1HHHHIU8Dff.ff.LLLLHH9u싅HMM LAEhH5w9HHLH;waAH;LaDH;=aHP4HPAx HEelM} AEhL}4LaHIEHEIuHHHLHHH1IH?HH!I)M~SHJA2@/ fρLHH9u〽@t A2E1A>IHtSH+8L`IA$A$LLPy;L`LPfD{5HLc`IH^LA1HNH5H:H 4HXLeH;Ht xtHI9u':fL牵P:P^E|$AD$II {f.H51HPL8-1HL8=wI}Px HӍIEPIcE0Iu@IM8H1L L HH9ufL9L^L8RL9!HH)HH9HDOGII fHA]HhH5H813K@E|$AD$HLML8II IuA$I߅A$LHP9HPvfDH\H5H8:0H5H=1LHP8HPDH5H=R1諪LL8M^8IT$Lj]L8fD>6I>6IEcH[H5H8:9PLHP7HPDLHP7HPDH7H5H=1軩3L\IDOGII IL牵PI7Pt[A$t4A$yݾ6GLP6LPGH6A$S[ LP6LP`A$A$Ll6A$XA$KLA6>HPHAH;Z=wHPH=Hv1IHx H=|1LA$x A$ H5H=#1ħHH=LHYH5ԢH86LLMHYZH5ʢH8j6H 5BL5[3@/fHJLH9t LRHH9u/@HXHJH5LHH81U.H;Yu7H{YHPPXHHpHLML8H;XuHPPXHH5D{HP2Hff.UHSH(HGdH%(H]HH1HUHuH}1xHC`HH{H{83xH}HUHu/H{PHtHCPx H{XHtHCXxt]HEdH+%(HCHH]H@DShcH{HVClu1l-E2f2kfDK0HS@E1Hs8 VH{@u7HH9P0H>,HEdH+%(uH]ÐHh-HC0DUHATISHH.HHtHH[A\]fHEHeWH8-,HUt`2HLH[A\]Mfff.UHAWAVIAUATSHhdL<%(L}I 10IMHL-V~ #fHnH-fHnHbMl$flID$AE=PMl$ AUPMl$(DAEfHEID$HI^)U)M)EMOL,H>HuNHUHIN wHLL1HMLuHUIARILLxşA[A_tiHLx~f.HHI<uH THLAL /H H8SH5Λ1)AYAZLmI>Htx IM9uHU[H=A$x A$E1MHt HIV0=wHUI~(=wI^ H}=w H}H]Luƒ>H}E1Ht'H;=SAH;=SDu L9GE=wI|$x 'I\$A$HvI9D$L9D_uE$EIDŽ$E1LmI>Ht xtNIM9uAHEdH+%(HeL[A\A]A^A_]fDMl$ Ml$(+.fHH AHQHH5L HLuH8S1'^_x%xA'xHf.HVIN0wHMIN(wHMK-PfDIt$@H߉x'I|$HxKAEMl$H=7AE+fID$hE18O!E1xAx,xfHPH5s18IfH]H}-HLaE11ARLuHUILLLx葛ZLxY'f.H AfDL,Kx&xHfA0H=or11uH)H=Ax _*Ml$(`fff.UHHt3HqH xPHPHpwHǀ]fUHSH(HGdH%(H]HHH"HUHuH}(xH=OH9{H9{H2H{0HtZ5 Hc1H?fff.HH9H;(IHLH! IHMx IM9wI>o(HHAEx AEtzLIf.HuE1LvpHcNdMHPHCDH81QHfHIyH9HXHHqH~%1@ff.H;T HH9uHAHJH5ZLMHWLEH81LELMnHLLEaLEIDLEHDUHLLE,LEI9HHH9H5H8Hff.HH9VHuH;,?D=@UHAUATI1SH%HMl$pHIcD$dMdM9sXI}HHHCH;C }n w HKHHHCx IM9rHeHxiubHHESHEPDHHHUHU؅txttxtWH9H=41H[A\A]]fH]xÃuHfHfDHHUHUw|qfDUHAUATSHHx|I1HHMl$xIcD$dMdM9sXI}QHHHCH;C  w HKHHHCxtKIM9rHHHHEHEHfDxOuHH{>fHHHUHU؅]xt8xtCAHH=1H[A\A]]fDHHUHUfDHfDH5ybH=^1?떅xu9ff.UHAUATISHH1HHTM$IcD$dMdM9sXI}{HHHCH;C  w HKHHHCxt=IM9rH H{xdu]HHEHEK@HfDH)_HcwdHCH@hHH@HHH H[A\A]]fDxSuLHkBfHHHUHU؅#x xt'HHH=H1[A\A]]fHfDHIHtXHHA$/A$"LHEHE HHUHUWF\ODfDUHATISHHGH5pHHHHt}I|$X?IHHH(Ht1x A$x A$taH[A\]x A$x A$HTH=YdH1[A\]fLHEtHEH[A\]HHETHEYCH x{Hnf.LWH2UHSH(HGdH%(H]HHH HUHuH}HxHH;8t#HtP8HǃHǃxH}HUHusHpHtHǃpxt/HHEdH+%(HH]fH HSHH9B0HS HEdH+%(umH]fDu:HHHǃ YHEdH+%(urH=ˁ:1f.UHSHH苐H7HpHp=wHt xtFHH9tHtP8Hǃ~+HǃH]1H6fu+HHtHǃxuU볍rH=ـ1fff.UHAWAVAUATISHH8dL4%(LuIHDžpHDžhHDž`HDžXHDžP=wA=wAL H(HffInI|$HWfofl)H9t;HXHHqHZ1fDHH9CH;T uHH5cLH +HH*HCH5nHHH+HHpH+xuHH@H@H; '4H; 3r H; 4e HH@fH@%,x pHDžp-L-ZgH=VIULwHHT.=wL-.gH=UIULCHHf0=wHAH@HH5\fHH1H@ILPMr1x R%HH3H I9Gu3foHuLH)]yLH`Iŋx %HDžPM/H H9CV5fInfHuHflLmIH)E,yH0HpAEx AE'HDž`Ax A(H0M4Ax AJ)I|$HTHDžpH9YHH5dLHM,ILpMq#H5kL9$IEH;20IUAEuHHx AE'HDžpHTI|$H9t6HXH HqH 1HH9 H;T uHH5hLH,ILhM,H5jL9&IAH;21l0AYAx A)'HDžhH0dH=RHDžpHSH: IH8=wAI@LPLL(H5[hHH{;L(H@H@HXH<Ax A H*0H@HDžPH H9C<H@HufoH~i0)EH)mSvH(Hhx HDžXH(B=A$x A$HbH=QHDžhHSH IH>=wAEIELpLH5@]HH @ILPM?AExAEuLL@ L@H(H5RHWHBpHH@HL@L@HHpHAH I9@BHuLHHEH]LtLH@Hhx M0HDžpAEx AE:0HDžPH@BH8L%N\HDžhLLMuL HH@H@HH/HLLHH@H8L%[LLMuL HH9H@HH0LLHpHH9H@H; DLwLoA=wAAE=wAEx :HuLLuHENsHhIAx A<AEx AE~7HDžpM(9A$x A$M:HDžhHL`xM<$Mt L;=,z)Md$MuHDž E1H(H-EHHH8LLHHHH8H8HIH9HRHhIHk0%HpIHN0L` L%+A$=wA$H(Mf(HWHBpHV<H@HI<LILhM/Ax A:H@LLHDžp/Ax A<HqHhIHE8DHpHH(8Lp A$=wA$H(Lb(HOHApH^BH@HQBHHHILhM7x ?LHpH7HIH}7HHP A$=wA$Me(H(LLHDžpHHAEx AE`CAx AZCH HhIHQDIH;DLp A$=wA$Me(H@LH(HDžhIAEx AEFH HLLH tH x QMtAx A QMtA$x A$PHCL-1LLMOH=rO1LHAH8VL8M"Mx 6MAx A/OL0L(L=E(Hf.ff.HH9dHuH;&Rff.HQH=IHSHIH"=wAELLLpÃ]$AEx AE HDžp0H>`H='IHSHIH=5=wAI@LPLL@H5_HH6L@HHXH6AxAuLI|$H5=YHGHH9ILPM)I@H;%A=#L`AL@x A')H@HY~N_HEfHnG fl)E@u tEH@H}H@HHNwHPIH)H@x )Hd%E1AH9C;H`LMLLEL@HEHEL@LHH`I:0H[H ^HP =wLHMHL)H?JtLH LL@L@LHpIMtAx A[4HDžhAx Aa.HDžPAEx AE'.HDž`x .HDžXM@/Ax A /HDžpH5]H0H9t-HCH;#-HCHH@HӃ,H8HRLHMuLH@H.H@HH?H@LLH@H.H8HpQHDž`LHMuLOHHI!H@HHcLLHXHH7!H@H;"1LwLoA=wAAE=wAEx uHuLLuHEhHpHAxAuLfAEx AE,HDžXHz xuHBHDžp"HLpxM>Mt L;=)"MvMuH(E1HDž HHSH8L(LHH8H8Hq11LHI?uHpIH11HLuHXIHE1E11LHLAx A/E1E11LLLHDžp訆)#Ax AHDžXH(LL(MtAx A CMtAx AEH Htx DH@L-CH@HHDDH=iA1H@LH!HWAH@x pDHL= A=fE1E11E1HDž@E1A1MtAEx AETMtAx AMtAx AHtx Htx MtAx ADHE1H=yH@tH@xtYA$x A$tXAxAt)HEdH+%(sAHeL[A\A]A^A_]@LfDHfDLfDHL8qL8DLX#LLL(L0H8,LL(L0H8gLL(L0H8L(L0H8ELL0H8L0H83HL0H8L0H8HPMML(LHtx+MtAx A!HhHDžpHtx %HDžhMtAEx AE H3H=FwHDžX莧HH`HpHX={L`Lp1HXLLL8HL8HHPI-H@H@HH-H=eL(L8HL8L(V/1LLH@H8H8L(LH0H@H@CAx ADHDžPLhE1HH;H;H;^HL8H@H@L8x 05[1Htx *MtAEx AE*+MtAx A-+HH@xH8L8Htx +MtAx A+H HBfL=qA=wAL0DDHf.ff.HH9HuH; ff.L0"HA}x AE#ID$H57PLHH%ILpM L;-L;-C L;-|6 LFÅ AEx AEHDžp@L=5M9BIT$ID$H5sPLHHH ILpM 111LflHHj AEx AE; HDžpHCHCHHHH)HH] H: HHH*x ID$H5ELHHHHHCH5[HHHHHHpHx HA9HAHƃHHH)HHHHH@H@HHx fL-NJH=8HDžpIULXIH=wAIBL`LL@H5DHHL@ILhMAx ABHL@L@HH`IL-IH=78LH@IULL@LHI=wAI@LPLLH5EHLL@HL@LLILXMAx A H I9AHEHELUL@HPHH<L@1H5BHq wHƺH}LmLhH)H?HH H4LLHHHLH@Hptx (Ax A HDž`AEx AE HDžXx M HDžPAx A< H@HDžhH:HDžpL9H@H5ALLpID$HHaILpMuIEH5"CLHH!ILhMEAEx AEZL; HDžpL; NM9ELLFLA"Ax A*H8H5SAHDžhHHHSE}HHIH,H@HHHCLHIH8H8H@LLHIUHHLHH 'H@HHHF LLLHHPH&H@H; /HOLo=wAE=wAEx #HLLHMH HEXH LHhHËx (AEx AE!HDžPH5&x A$HL H8HDžhAVHHH(LL93L YL-z4^IBHH/H=VZL8RL8(1LLIML8'Ax AOAEx AE L0A=WALHQhLDL7L*oHBhHQ+HxF+H(L@eL@HH;0LLPXLH@H@H`H,AK@SH)HHCQH)HHLbHU-L0E1E11HDž@E1AhDCSHH HCSHH HHHH@H@HL0E1E11HDž@E1AAQHH HH*AQHH HLULL@AL@H-H AEGOLH(HXA=wAMwA=YLAH H(AALfHHHjIH'H@HHHLHIH0H8H!;LLHIUHHLHH| H@HHH_LLLHHpH[ H@H; *HOLo=wAE=wAEx HLLHMH HERH LHhHËx O"AEx AE9HDžpHx AL HDžhHH8AVLHLHHH(-XHZL L-.IBHH)H=TL8L8u71LLIML8p,L8H2Ax A+L0E1E11E1E1A LɅHL8@@L8|LuAALQLDH,I|$HDžpaE1E1E1E1HDž@AFHLLIL,M$IT$I9HXH'HqH~1fL;T vHH9uH=IJ1E1HRH5UAH81E1E1E1HDž@L0qL(H5Axt3LpLhHDž@L`HXLPHH8LpHDž@LhL`HXLPH8LLL@qLL@H@=IE1E11HDž@E1A\HLLHHBLHHGLPLhHH@xH8L8Htx MtAx AH tH x L0A0HDž@L0E1E1E1HDž@E1AWH %HL0A1HHHwHXHz(=wH1HuHHHEHELHpI‹x M!L11L@lSL@Ax AE1E11E1HDžpE1AHDž@3IL0AHH='HxLHxHH+E1E1E1E1HDž@A@HH@ɧH@HH@ZH@HL0IE1E1HDž@1E1A?CH=&HxH LxM3L`H@H'HDžpLPE11L0A#HDž@DsILpLP1HDž@L`L0AwH=&HxLLLxML@ E1L@HH5)L01E1E1HDž@AHPL(LLXHrh]SH=]%HxLHxH}qH%A1+LL,LHDž@LPE11L`L0A#0A=wAIH =wLIeL`L@[L@IHLHLm1E1L0E11E1AL01E1E1HDž@AfA H@I_LH3HH8HDžXH@LPLhxH@KLpE1HDž@A0HXL0H;LIHH;H;YL;5QLLAx A#TAEx AE6#HDžp+I|$H"#DH;LϺL(_L(HH@O H@H;= H;=H;=L@L(LBL(Ax A["LpLP1HDž@L`L0A:MGIOA=wA=wAx ApHHuLEL0H@LeEL0H@H`IA7A+LcH@LL(H8=LL(H8LL@`H= HxL*LxL@LML(L8L@L8HL(H]HLH5a{LL(E1AH81L8E1L0L8L@L(L@LpE1E11HDž@E1AXHL=+!HPML(LHE1cHKL{=wA=wAx HuLHMH@LuLmCH@H0HplaH'TLL01AHDž@6LLL@IwHpVL`L0E1E1HDž@A&LLLH Hf.LHbLL8H@-L8H@LAx AHDžPIYIQ=w=wAHhx ALUL@HEH]L@HHPH$L0ATHLALHpLhHtx t HDžpMtAx A[ H`HDžhHtx  DHyH=AUHDž`艅HHpHhHP8YLhLP1LpLLLLL8L8LHHXI=HCLHH)H=CL8HL8L1LLHLH8H8LLHi DAx A HDžXHpH;H;H;gHLLH8H8LLx bMtAx AMtAx AMtAEx AEHH@xH8L8Htx uH Htx mMA$A$L<م.#HLL8 L8L7H=HxHLxML@L`L@H( E1E1E11LPAE1L@L0LhH0ff.@zfH<2HDžX=wHHuHHHEHER=H@Hpx LPLhHDž`H@`H@11L(L8HCL8L(x lL0E11E1HDžpA.HDž@LDH7L*L]H;IH0H@HH@H;=\H;=2 H;= L@LAx AnLpLPHDž@A-HXL`LhL0LFuA=A#L(H@}LE11E1L0E1E1A%HDž@HHH8LpLPA0HXL`LhL0IA=ALMH@HpI9+HXHnHJH~1L;T HH9uHIJ1E1HVH5-@AH81E1E1E1L0L`E11E1L0AI HLH8L`x LpLhL01E1L(ALkH[AE=wAE=wH@x jHuHLmHe,LeHEH0HE-9H(HhAEAELHLAL@L`1E1E1LhL0HDž@ALND$L7L8&L8ISwHP=H=HxHLxML@L`L@HE1E1L01LPE1E1A%L@fDwHpHLLH88LLH8LXfHuH)ELI7HpHj5H=HxHLxMJL`Hp1L0E11H@L(E1AL0E11E1E1AL[LL@GL@L`L0E1E1HDž@A%Q L@HHL@L@LLL L CLL L LL@L@L|rLeHDž@L`E11L0L(Ap+IؿHBhH(LHHxDILAۅALE1H@1LL(L8L8L(HH@H@ۋ)L(GL(Ax AHDžXLH LH LLLH8LLH8+LIHL ~L LH _LH HLH8JL`E1E1E1L0L(AQHL L L`E1E1HDž@L0L(AHL@L@RH@1ҋAx AHDžPE1E1HHL8H8eL8Ax A LpE11E1LhL0ASM`fHnMhA$fInfl=wA$AE=wAEAx A HuL)E2H@HhA$!A$LL`1E1E1LhL0AL(HHL8H8EL8Ax A LPE11E1LhL0A 3LL(H8L(HL=H@HD1tfA.EEqLwPLKLkALh=wAAE=wAELXx  LE1H׉@L8@H@ۋ1ffA.AE鄸HL L L@rML@Ht M;U(HzpLjpHtx Htx MtAx ALPE11E1HDžXLhHDžpHDž`3 L8HuHiH5j2H8 L8Ax An L0E1E11E1E1A HL L LpfHuH)EL.HhI`HAhHHlHxaH(H=HI鈽L(L8L@L(L8L@9L(L8L@L@L8L(?LL(L8L@L(L8L@H@L(A0L8ELpE1HDž@HXL0L8L(OHDžpAHHXHH@xH8L8Htx EH Htx NMtA$x A$L0L(E1ALc铼LV陼LL0H@;L0H@gL WLd LL8L8 H"HH8HH@H@6H<H5Uc1E1L8AH8L0E1E1HDž@L8HHH8acHLA L1҅|Ax AHDžX11LHL8'L8LH‹(.H(L@$>L@HеLLL8菿L8LE11E1E1HDž@ALL8(L8LHMt M;i( H{pLKpHtx g MtAx Ag MtAEx AEZ HXE1E1E1HDžPHDžhHDžpHLL8茾LL8uLL8fLL8LL8@LL8L0E1E1E1A&R-H8LHL ALL8H@ʽH@HDžPL8CLL@蝽L@HHI9oHuL;z]LX$LPfHuHL L)Ew(L HhHLpfHuHL L)E5(L HhHLL1L8L8IAx AjML0E1E11E1A LL1L8蜻L8IAx AML0E1E11E1AYfH@L(L8uHHLMALArAfL觻Y蝻HL81A聻LpE1LhL0L(L89ILpLP1AL`L0`LHLL@L@LL)@ۺfo@LƺMHLA LLLLH8聺H8HDžXLLL0E11E1A.邿L:L-H@H;H,H5 LLķLH@HH@H@LpLPL`LhHX1L0A0H@鲾H@H@LpLPL`LhHXH@LA0L(L81L0L8H@L(LH@ϸL(L軸L许HLH5bZH81>AL聸fLLL8fL8LFζLpLPA-HXL`LhL0YHrH1H5YL8A#H81蟱HpE11LPL0L8 HHI9HuL;H(H5AZ1E1L8AH8跸L0E1E1L8鞼L0E11E1E1A遼L9L,L@1LLSHAxH@Hλ1LpLPA0H@HXL`LhL0H@褶H藶L芶ۺHH5%H8请H@H8L(L_L8HMx)u"E1HL8L8MLpL`1ALhL0L(LL@L@1HLE1H5kWL8AH81:E1E1E1L@L0L8駺LLH8EH8L@L@6L@HL@L@LL0E11E1E1A %H>LH5VH81z7H(H"3HI'L8藴L8L胴LvLLVLL8+1Lp1LhH@L`AHXLPL0UL İx LpL`qHkH5l"H8 L81LHImL0鐾HI|$騦HH5!H8虴L8QH@AEHH5!H8fH@S"HH5!H8:H8LL7ѲޮLIJL跲H HH5kTL8L@H819L8L@HHH5/TL8L@H81L8L@HHH5SL@H81ȫL@bH@LA0L(L8E1L0L8L@L(L1m1LpLPA0H@HXL`LhL0鵶f.f.HGPw HGP@HG`w HG`@H-=wHHGhHtwHHGPw HGP@UHAUIATISHHHpHtHAԅH{ HtLAԅH{@HtLAԅH{XHtLAԅH{`Ht LAԅuwH{8Ht LAԅudHHt LAԅuNHHt LAԅu8HHt LAԅu"H{x1HtHLL[A\A]]DH[A\A]]DHHt4=wHzXHrXHt xt1fH5qUHG1]UHSHHHtYHH HwHH(HwxtH]1HخfDH]DHHt=wHfDHt>UHHHUH}9HUtH}H=wHfHaHHt=wHfDHt>UHHHUH}HUtH}H=wHfHUHSHHHpHtHCpx H{ HtHC x H{@HtHC@x H{HHtHCHx H{PHtHCPx H{XHtHCXx H{`HtHC`x H{hHtHChx pH{8HC8Htx ^HHtHǃx FHHtHǃx .HHtHǃx HHtHǃx H{xHtHCxxtH]1諫H]1蛫fD苫fD{,fDk>fD[PfDKbfD;tfD+fDfD fDfDfD۪fDUHHtSHFtFH=wHzHHrHHt xt 1]{1H1H53H8蚫]UHHtSHFtFH=wHzPHrPHt xt 1]1HH5r3H8]UHHtSHF tfH=wHz@Hr@Ht xt 1]{1H1H5"3H8蚪]HH523H8zUH;5<HHtLHtGHV tRwHHHt xt1]1@è1HyH52H8]ff.HG@Ht#wHG@UHHH}oH}HG@Htw HG@fff.HHtwUHSHHHtHwH]fDUHATISHtmH;5HuqHH5n2H8֤=wI$I$Ht xt1[A\]CHQHFuHH51H8MfDUHATISHtmH;5HuqHH5>2H8=wI$I$Ht xt1[A\]胦HHF uH$H51H8荧fDHGHHt#wHGHUHHHGH}H8HHMHAHHtw HAHfDUHHSHUH(dH%(H]HH=ԣHMHtHEdH+%(u?H]HHM藟HMHuHHH5'GH81HMȣUHAWAVIAUIATSHHLHMLEIIEHM0ff.ff.ID$IHHL9xuL@M;FuDH AN Dʉ@@8uA 8Hx8 Iv8ICuHEI)IL H[A\A]A^A_]DHpI;vu>D@ AN D‰@@8tUff.ff.HI9tHHL9xuf1H[A\A]A^A_]A uHx8 IF8HHruH_HULH5/H81/IF(Iv8@HE@Hx(H8A@HDHx(H8A@HDp IH MF(IF8@IEUfDUHAWAVAUATSHHGLEHHIIIIHuEIFIHt7H8H賣tnuٸH[A\A]A^A_]DIM9I$HH8otۃuHHUHH5.H81DM)IM7H[A\A]A^A_]HHUH5~.H81褛V1MHHt=wHfDUHAVISHH輣HH=wHA1E11HMHH=蔝HMH‹x HHBHUHHHH HUHx HtnItGxtI=wHH[A^]fDH =w=wIfDӠH =w;ufHXtHHUDHUIKHHMHMHUHH=7H逗HGXHHt wHGX@HGHxHt3UHHHM2HMHAXHtw HAXHwfHtf=wHff.UHSHHۖH{(tHH4HH]闖HwPH1H=,GחUIIHHH HGLP@tluwHHMDHMHff.UHHHH?HHt xt9Ht xtGHt xtH|HMHu|HMHufHHMt|HMff.UHAWAVAUATSHH_pHGpHLcA$=wA$IHHuII;HuHtwA$wA$wL&IIIUxL*HA$x A$t]Ht xtcMtAEx AEtH[A\A]A^A_]fDHL[A\A]A^A_]Z{f.LHED{HEfDH0{fDHHHHGxL(HeUHAUAATISHH}~Ht[H5HHEH}LvxtH[A\A]]HHEzHEH[A\A]]DH1[A\A]]UHAVAUATSHdL4%(LuI(yHH5HUHIYxH]Ht-H5HUH=xLmMxt+>zHEdH+%(?HL[A\A]A^]HyfDsHtyHEdH+%(HL11[A\A]A^]~fDH I9L;-ĝ@@u L;-luRI9tYxt0AEAAE4Ly'f.HyfDLptAEx AEt=xt&A$!A$LxHxLxwLVIIM1DHI9tM9D uf.1ff.IT HBt~@tuL9tIXHt,HqH~[1HH9tGH;T ufDLDHH9pHuH;̛^ff.HI9c1HHH9tzHFHHW@~tw@tnHXHt:HJH~Q1fff.HH9t7H;t uf.HH9tHu1H;5f1D#oHNH~1 HH9tH;| ufDE1HJt H9UHH HGtf@t]HFH@LXMMQ1MFfDHI9t7I;t uɸHULEHMH}TnHULEHMH}uIL9tJt H9tL1étHULEHMH} H}HMLEHUHHH9kHuH;5uXUIHH uuHGH;thHHpHHyH}HHUHMDvHUHMHH@H}HEQHux fDHHy AEtI;Jsk=wIBHt?1 }1H9})HcHATD9~މ9|A9@AQ1A9ff.@swHӳ@HcH>DHHHHHHHkÅH HHEÅHHHEHmH[HH8H"HHlHMH=ÅHHHEHHf1Ht)H9HOHVDG\H9t&AHt1~\Ht1H9D8F\uF]8G]uHcGX;FXu…~BH1Dff.HH9tHLH9Lt1øAS{V`19W`mLGHNMHSI81HH4HtHDI9DuUHH fDHDI9Du7LEHUHMt$HULEHMHI<HtH4Hu11H1H<1H<fH~?UHAVIAUIATIS1ېI<$AT$AHI9u[A\A]A^]H~OUHAVIAUIATIS1ېI>AVf*YADHI9u[A\A]A^]DUHHHH?MEPMYEXfff.UH?IIHwHt/u+HHI9LBMuMHF]1@Hu HDHytHHH H5H81Wa'DHaHH H5H81.a1]f.H1HH H5MH81`@HyDoUH?LOIHt2u.HHH>LBHIuKH6IA]Hu HDHytHIH H5H81W`'DHaIH 0H5H81.`1]f.H1IH H5MH81_@HyAoG<4wH2HcH>DUH@H5H81H_1]@ff.f.f.f.UH HLGw@DLHMtbIH2H6H9t,IHLJLH]H5fH LH1^fDHSIHILH51]^fDH,H5^H?IHOIЃtHAHLLHtIH>HAHLL@UHhH5 HH H81H2^1]ff.H?ILOHуtLWH8HIAHLHtLHHfDUH؇H zH5 IH81H]1]ff.wD@UHAWAVAUATSHHHGHH Hy*@s@pGE1HcQX@hA~KH<уAt)LyHH9t/fDff.ff.LyHLyH9uCGS@HC0@QNEIHu$4EHff.L(CFMgMu<@f<^F<4H=zHcH>fS@A@QY@>~3NEUH &uEP@PrEO@OdH@H5H81[HCS@EsD4@H@H5H81[CF1<@AF\I9vt;fDfDHH5H8aCF1<@|f{DHuHH|LS 1HuLHHt I:H)HC H{86{DHuHuHC8HHH HHH@fDfDHHH HHHHHHDI~HSIEHrHBHsINHJHSHB>@HCHPHSLhI9tH{f.II9uHC0HCH CD1C@H[A\A]A^A_]I~`1yXLC0@@{GHH HQI9uWAfDfDHH5RH81X1ÃɃCMjHLH5yH81W_HDH5H81W?f.UHAWAVIAUATISHbHAHcH>AD$E=AFIAVfW!,MfWADHI9u[A\A]A^]H~gUHAVIAUE1ATISHI>AVf*YWZLZWCIL9u[A\A]A^]@ff.UHSHHEM@ff/wFH;Sf/`r ^\\f(OMH]YE\DXNYEH]XEUHAVfI~SHHM@H;Sf(\f/vNfWNMH[YfInA^]\@ff.UHSHHEMfDH;Sff/v \^NYEH]XEff.f(f(UH\\HH0H?f(U^]MEu}P}]Uf(f/r,Y]eYff.wIf(QXf(ff(\E ?\YYff.w'Q\f(f(e=Uef(U(UU̐UHAVSHH EfWIEH;Sf/EH;ESYEQSMf(fWY]f/rqf(LfI~ELf(fIn^XTSH,HwMff.E\H H[A^]f/Mr fff/<r @UHAVSHH \f(ESMfW ESEH;SrH;f(E\YXES B^MfI~EVS-f(2fTf.v;H,f=H*f(fT\ fUf(fVf/O=f/=f(MU^XRUfInuf(\Y^Yf(\o^f/H H,[A^]UHAVIAUATASHH ILخIM5DL9HHIE;butI=wHL1HEIHD`(HOx AEx AEH [A\A]A^]@MEpIEpMMHA=wAIH(H=HDLLMLEHMKHMLEHLMHI;H(I}pMEpHtx Ax AHtx L^M5GDLTHc9YHcHLD;a9EHcHM)HwHPHHHHHHHHLLKHMDaH=##fDeH߉W fDLH [A\A]A^]M@HMHDLHMLMLEILELMHHMHAAAx A@HHωYfD֫9H@L׉UHcMHEIHM؋5H HcMHHH9kDHDLHHHI}pIEpH_Lf.HDLHMLMLEVHLELMHHMH8AAALHMKHMHtLHMKHMHMLMKHMLMfDMHHHGH8D`H=HI =AAALHMKHMfHLLMHMLEJLMHMLEH9H=wJLHMLEJHMLE1AAh@UHAWIAVAUIATISHXH;HAdH%(H]HH9t6HXHHq1HfHH9{H;T uL5;HL6HL1Ҿ=8HH*Hd<H fInH}ID$fHnI<$HshH@@flLeI)EfoHE)EfHE)EPHHSX2H{P~xHKpAEH1H2HKxHH9H9u t%fHHHKxH1MLHHUdH+%(He[A\A]A^A_]HH9dHuH;LlL5]9iFfKdE1HkH5H81A|fH tH9HHHHH9lHHt H8H7k1H5H81LAMtA$x A$tF@L58@HjH5MH81@IfDL8GfDHoHjH5H8DHcH~rHHH5ijHAL%H>PH51n@XZH;HH%jH5&H8GfH HqHiH5RH8GHi1H5uH81?H9y/rkHi1H5H81?eHi1H5H81?F]Dfff.UHAWIAVHAUIATSHHLDdH%(HE1HHIEHHc1#ff.ff.HH9 IՐxHhH5H81>E11HLfoAfoAGfoAG fo AG0fo0AG@fo@AGPfoPAG`fo`AGpfopAfoEAfoEAfoEAfoEAHEdH+%(HeL[A\A]A^A_]DHDIHHE1KD4 IL9tMK|5EHuA$A$LCHcDIHHH@hHH8fH=w HHDIH~H;@HHCHHH!A$=wA$La Lq(HQ0=wHY8H=׊H1H6HIMx A=wƒ+xuHLyBLIEULLH0LH8HHL~HLfoHL$fo$fo$fo $fo0$fo@$ foP$0fo`$@fop$PfoE$`foE$pfoE$foE$AoE$AoED$AoE D$ AoE0D$0AoE@D$@AoEPD$PAoE`D$`AoEpD$pAo$Ao$Ao$Ao$Ao$UiHĠLA$x A$t$A;A/L2@"LL@LLL@Lf~HL?L#Hv=RH]Ax  H H=YHl?|Ht[xuHLF?LA$yE11DA$uLL?LE11AxAtdxuH>HAH=H*AxЃAuL>H>LH{>HA<DUHSHHHHPHXL`Lht#)p)M)U)])e)m)u)}dH%(H81HELPHXH@LZHpDžPDžT0H`.7H8dH+%(uHpH=~6;@H;=aH;=au H;=ebu +5ff.UHSHHHGHHt/HxaHQH81">u)HH]DH`HH5H81u6xt1HH]DH1HM\EAV%TfZUAYAf(Ykf(fA*YYEXEZf/wFI>AVfAA A*AYDÈt W{EE;<<EH[A\A]A^A_]I>AVf*YW/ I>YMAVf*YýWMW(XY/vX AMQW M@ff.H~7UHAVIAUIATIS1ېL"AHI9u[A\A]A^]@ff..(zbf.zUHAVSHH =M/=ü}f.H;SfH*YUU$mU fA~\/i u(^fAn/rH [A^]fu\5C?=/YuQ^}=}H8!f(EYX/s(MH;YYESMûf(*YYEYY\/wXMEE5f\u(M(XZYEYYX/E+EYEH [A^] M\^EEuME(Y\( Ժ^fAn\M/H [A^]"UHH?HP]Hf.UHH?HP]ff.UHH?HP]Hf.HHGHf.$E„Qf.Yf(D„7UHATSH -f/fE1XH@q HP^Yff.YHXHHH9uf(MU]UMf(5Pf/u^f(\YX 7X\vGMtBI\MHf(U!MI9U\uH f([A\]Ðf(f\L,I*Dff(UHAWAVAUATSHHHf/Esyuff.zAuHHL[A\A]A^A_]EfWsE1  CE@IMH;SMYf/Mw+EEQYXf(EYf(\ \=f(\-^ f(mXfI~XfI~^\fI~@H;SH;f(\USU fInEEf(fT;\^MXYXEXQMf/ GL,fInf/ujMcEMEfInMmEEY^XuXuID$f(f(\fI*YM\MEfH*M M\f/MM%wf/Bmf/3Uf(HHH}\^f(f(H}Df.f(ff.zUHH f/vX\MH}f(H}EMf(ff.wrQXf(YXEf.Y pH}Uf(6fH}HH*XEKf(?f(UUf(vUHH Mf(H}EeH}f(Yf(M]MY^f(fDUHHM YEUfH*HAWAVIAUIATISHPt H9r  PMfAf(Aff(\M]Y\f(ef(fmAFX]AV(f(U]Y]H,fUMf.I^0f(h Qf(Y]Y%<-\f(fTf.~5Nf(fH*f(XuXXPfD(f(mf(DXfH\A^8DXf(efAA^H^f(\XfD(]A^X]Yf(f(\^f(YXYfA(\UAYpA^`^Yf(XYfA(AXX`AVhYf(EAFpfA(^D^Xf(xAFxfA(XEA@I}AUMI}YMAUMf/Mf(f/Muf(mfH* D\Y^XXE\XUfT.^\f/l]M]ML,MI)LHIHH~$hYEfH*\f/dID$fe^eH*YI9?f(f/PM)f/eMGH[A\LA]A^A_]fff.ADEf/xwif(M]>^pXEL,M^EUM\M]Y YpDf(M]X^`\f((L,M9E]M\xY cY`hfDHCI9IVf(@ff.ff.ff.ff.ff.ff(H*^H\YH9uTfIFH9>HSff.ff.ff.ff.ff.ff(H*^H\^H9uff(^ HfX hHHf(YXX ^H*^^f(XMYMUUMf(f(\f/JXf/IFfEfL*HCfEfH*f(ID$fE(A^H)f((L*YLfE(EYL)D8EYHD@H*f(}YDD  D@ED0fA(^E uYuUY8@f(^ -BLfD(:H)Df(DM54fD(fA(fD(DD^ DYH\f(f(^D\D^A\f(^ fD(D\fD^(H*XED^Y@AXfEM*EYfD(AXfE(E^AXfD(E\fD(E^E\fD(E^E\E^E\D^0D^AXfE(E^D^E\fD(E^A\^E\fD(E^\^E\fD(E^\^E\D^8D^\^M^DXDXfA/-MUYH\X2L,H,ffUH*f(fT\fVf(Rf.BWQj8rz HZ0mj@Yu5iHjH}mjPYuXjXmj`pjhh`jpmjxxmf( hMUMU fUHAVIAUIATISH0E؅t H9r?UMff5AI*AV\]Av f(u ]Y]U]MEYAFfYAVXX f.QY Xf/H,I^0I}AU}1f/f(wZ@I}AUM1f/vAHH9|Lf\H)HH*YUYfH*YU^f/wH0[A\A]A^]Df.Bj ZHZ0m]Kf(UMEYUX@f(Hff.EunfH*f/rY f/rMUHH Hu\Yf/f(r"HHEH)Ð1D  HHEH)@ff.UHHHH?EPMf/v3f(ظf(\fYHXf/w@ɸÐUHAVSfH~fI~fHnfWYd<f(fInfW3^f(fTf.v3H,f(f%fUH*fTXfVf(f/ֺs[H,A^]f.H[A^]Ðf/$r D 1HUHAUIATISHHH HHHH HHH HHH HHH HH H HH u*f.I}AU!I9rH[A\A]]fDI}AUH!I9sI}AUH!I9rfDUHAVAUATSHHHHtlI̹IIH?H9rlIEH9EsuDe؉ID9s!1A9sI}AUI9rH IHH[A\A]A^]fDIEHEuULcIHHL9s2H1HIHH9sfDI}AUHLHHH9rI>닐I}IEL!H9rIHH[A\A]A^]f.I}IED!9rIfHuHU؉H&@HuHUHfDUHAVAUATSӉHt_HGIAH?AEu_suDe܉ID9s"1A9sI}AUI9rH AH[A\A]A^]IEI}D!9rHA[A\A]A^]@uЋUH[A\A]A^]@ff.UHAWAVAUATSHLefuALAAIfERUH?AVUA$A4$DAf9AAřAAf9r,fDI>AVA$A4$AfD9sgtA,$DI>AVA$A$D!fA9stA,$fDfEH[A\A]A^A_]DEH[A\A]A^A_]A,$AtA,$AEfA$@H?AVA$ff.UHAWAVAUATSHLeuALAAIEDrH?AWA$A$AD8AAAřA8r,fDI?AWA$A$AD8sgtA,$DI?AWA$A$D!A8stA,$PA$D!A8rEH[A\A]A^A_]EH[A\A]A^A_]A,$AtA,$AEA$DH?AWA$ff.tXUHHHAtHE(AAHEɃ@LMH?PHMLMAHEɃÐfff.HuOIHHL)ʃtI1IL9@ff.I1IIqL9uUHAWAVIAUIATMSHH(HMH9H9E'H}~~DzHED}EfI>AVMI9s-E19s @ff.I>AVI9rHEH LI4HHEH9EuH([A\A]A^A_]HEHHELzH]H}~I>AVHLHHHL9s-HE1IHH9s@I>AVHLHHH9rHUJ/IHHUH9UuWfH};1ff.I>AVLIHH9]uH1DI>AVLIHH9]uH}IHEII LHI LHI LHI LHI LH I fI>AVL!H9rHMLIHHMH9MuaH}SIHEII LHI LHI LHI LHA ǐff.I>AVD!9rHMLIHHMH9Mu@u@IHHL)ʃtA1IL9A1IAqL9uUHAWAVAAUATASHH(LMHMȃEDjAE1DmDeH}~eff.H;SUI9s.E1A9sfff.H;SID9rHEH DB IL9}uH([A\A]A^A_]fDH~E1H;SHuDBIL9euH([A\A]A^A_]fDH}~E1III LHI LHI LHI LHA H;SD!A9rHuDBIL9}uH([A\A]A^A_]ff.fuOIIH~AHL)ʃt#fA1II9t)ff.ff.fA1IfAqI9u@UHAWIAVAAUATASH(fEHBAL1fEDIIE1HEAĉEfff.E7I?AWAAf;MsJEAAf9s;Eu#I?AWAȉAfD9sAAfD9rHDfKH;]yH([A\A]A^A_]f.H~LIII LHI LHI LHA II1HE1ft%ff.D!fA9ȃuGI?AWD!fA9rDHfSH;]uH([A\A]A^A_]Af1҃1HDfSH;]tfHLM$II?AWAfHSI9HDfCI9uDDHfSH;]\ff.uH~@HLDUHAWIAVAUATASLHùEHGADrM, E1A1ɉEAƉEEI?AWAAD8sUEș}A@8sFEu.I?AWDD8DAD8sAAD8rẺHCL9EoArH~zM4 1III LHI LHA 1u#I?AWD!A8sKD!A8sD!A8rfUHSL9uH[A\A]A^A_]ffA%DH~M$ 11u1I?AWM̍HSI9tHKI9tuH KL9uoHUHAUATM$ SLHtKI11uI}HAUƒSI9tHKI9uH[A\A]]ÐHC@3L9t@sHL9uH[A\A]]@UHAWAVIAUIATSH(H}LMI~f IM`1@MA\ HL9tDHUH}LMA^]I)IDMH([A\A]A^A_]H~M`OtH([A\A]A^A_]@MMADVUHAWAVIAUATSHhHu1Hh[A\A]A^A_]@IH<LELHMHUIH2HuHtLHUHME1H?E1LHHEJ 8 F  ; W-+  Js Kxk vD  p( , F"w5 U l #   w ] ,4Y C1  H   f{^  ov4q  U &F  | Y|1bPz ~ dP1_ W   e \   e7 B    ;  y! _ `U, }P!cP3gg z 6@pqPAAa@z &N _#.>U@;0>  `N'ZpU PYf  IA$ Am@` 0T Pyd1p<P7% YVP#q0"pALP~`VzqYB (__gmon_start___ITM_deregisterTMCloneTable_ITM_registerTMCloneTable__cxa_finalize_Py_NoneStruct_Py_DeallocPyExc_BufferErrorPyErr_SetStringPyLong_FromLong_Py_FalseStruct_Py_TrueStruct__stack_chk_failPyObject_GetAttrPyExc_TypeErrorPyErr_FormatPyDict_SizePyTuple_NewPyObject_GetOptionalAttr_PyDict_GetItem_KnownHashPyNumber_InPlaceAddPyErr_ClearPyErr_OccurredPyExc_NameErrormemcpyPyList_TypePyTuple_TypePyLong_TypePyExc_StopIterationPyNumber_IndexPyLong_AsSsize_t__memcpy_chkPyExc_ZeroDivisionErrorPyObject_GetIterPyExc_OverflowErrorPyUnicode_FromOrdinalPyBaseObject_TypePyErr_GivenExceptionMatchesPyThreadState_GetUncheckedPyException_GetTracebackPyBytes_FromStringAndSizePyBytes_FromStringPyObject_VectorcallMethodstrlenPyException_SetTracebackPyBytes_TypePyNumber_AddPy_EnterRecursiveCallPy_LeaveRecursiveCallPySequence_TuplePyExc_SystemErrorPyObject_CallPyObject_ReprPyUnicode_TypePyObject_FormatPyLong_FromSsize_tPyNumber_InPlaceMultiplyPyGILState_EnsurePyUnicode_FormatPyGILState_ReleasePyNumber_RemainderPyObject_HasAttrWithErrorPyObject_GetItemPyLong_AsLongPySequence_ContainsPyList_NewPySlice_TypePyIndex_CheckPyList_AsTuplePyObject_SetItemPyFloat_TypePyExc_ValueErrorPyObject_IsTruePyExc_IndexErrorPyExc_UnboundLocalErrorPyMem_FreePyMem_MallocPyErr_NoMemorymemsetfreemallocPyImport_ImportModulePyExc_ModuleNotFoundErrorPyErr_ExceptionMatchesPyObject_GetAttrStringPyCapsule_TypePyExc_RuntimeErrorPyCapsule_GetPointerPyExc_ExceptionPyCapsule_IsValidPyObject_SetAttrPyMethod_TypePyList_AppendPyObject_RichComparePyBool_TypePyDict_NewPyDict_SetItemPySlice_NewPyFloat_AsDoublePyFloat_FromDoublePyObject_SizePyObject_VectorcallPyNumber_InPlaceTrueDividePyUnicode_ConcatPyNumber_SubtractPyNumber_FloorDivide_PyType_LookupPyEval_SaveThreadPyEval_RestoreThreadPyLong_FromUnsignedLongPyExc_AttributeErrorPyErr_SetObjectPyNumber_TrueDividePyLong_FromLongLongPyNumber_MultiplysqrtPyTuple_PackPyThreadState_GetPyNumber_MatrixMultiplyPySequence_ListPyLong_FromSize_tPyNumber_Absolute_PyLong_CopyPyNumber_NegativePyNumber_OrPyExc_NotImplementedErrorPyObject_IsInstancePyObject_GC_UnTrackPyObject_GC_IsFinalizedPyObject_CallFinalizerFromDeallocPyObject_MallocPyErr_FetchPyObject_FreePyErr_RestorePyObject_GenericGetAttrPyObject_GetBufferPyBuffer_ReleasePyThread_free_lockPyObject_GC_Track_Py_EllipsisObjectPyDict_SetItemStringPyInterpreterState_GetIDPyExc_ImportErrorPyModule_NewObjectPyModule_GetDictPyExc_RuntimeWarningPyErr_WarnExPyUnicode_InternFromStringPyDict_TypePyUnicode_DecodePyUnicode_FromStringAndSizePyObject_HashPyLong_FromStringmemcmpPyObject_RichCompareBoolPyType_ModifiedPyObject_HasAttrPyObject_CallMethodObjArgsPyType_ReadyPyGC_DisablePyGC_EnablePyCapsule_NewPyErr_WarnFormatPyImport_ImportModuleLevelObjectPyUnicode_FromStringstrrchrPyImport_AddModuleRefPyDict_GetItemRefPyType_FromMetaclassPyDict_SetDefaultRefPyMethod_NewPyObject_ClearWeakRefsPyObject_GC_DelPyUnicode_FromFormatPyTuple_GetSlicePyTuple_GetItemPyDict_NextPyInit__generatorPyModuleDef_InitPyDict_DelItemPyCFunction_TypePyObject_VectorcallDictPyArg_ValidateKeywordArgumentsPyObject_IsSubclassPyException_SetCausePyObject_CallObjectPyUnicode_NewPyUnicode_CopyCharactersPyImport_GetModulePyObject_SetAttrStringPy_VersionPyOS_snprintfPyEval_GetBuiltinsPyType_TypePyImport_GetModuleDictPyDict_GetItemStringPyThread_allocate_lockPyCMethod_New_PyDict_NewPresizedrandom_standard_uniform_fillrandom_standard_uniform_fill_frandom_uniformrandom_standard_exponential_inv_filllog1prandom_standard_exponential_inv_fill_frandom_laplacelograndom_gumbelrandom_logisticrandom_triangularrandom_logseriesexpm1floorrandom_zipfpow_PyObject_GC_NewPyFrame_NewPyTraceBack_HerePyCode_NewEmptymemmovePyMem_Realloc__vsnprintf_chk_Py_FatalErrorFuncPyDict_GetItemStringRefPyModule_GetNamePyCapsule_GetNamePyDict_SetDefaultPyBytes_AsStringPyUnstable_Code_NewWithPosOnlyArgsPyExc_DeprecationWarningPyLong_AsUnsignedLongrandom_standard_uniform_frandom_standard_uniformrandom_standard_exponentialexprandom_standard_exponential_fillrandom_exponentialrandom_paretorandom_weibullrandom_powerrandom_rayleighrandom_standard_exponential_fexpflog1pfrandom_standard_exponential_fill_frandom_standard_normalrandom_standard_normal_fillrandom_standard_gammarandom_gammarandom_betarandom_chisquarerandom_frandom_normalrandom_lognormalrandom_standard_cauchyrandom_standard_trandom_waldrandom_vonmisesacosfmodrandom_standard_normal_frandom_standard_normal_fill_frandom_standard_gamma_fpowflogfrandom_positive_int64random_positive_int32random_positive_intrandom_uintrandom_loggamrandom_poissonrandom_negative_binomialrandom_noncentral_chisquarerandom_noncentral_frandom_gamma_frandom_binomial_btperandom_binomial_inversionrandom_binomialrandom_geometric_searchrandom_geometric_inversionrandom_geometricrandom_intervalrandom_bounded_uint64random_buffered_bounded_uint32random_buffered_bounded_uint16random_buffered_bounded_uint8random_buffered_bounded_boolrandom_bounded_uint64_fillrandom_bounded_uint32_fillrandom_bounded_uint16_fillrandom_bounded_uint8_fillrandom_bounded_bool_fillrandom_multinomialrandom_multivariate_hypergeometric_countlogfactorialrandom_hypergeometricrandom_multivariate_hypergeometric_marginalslibm.so.6libc.so.6GLIBC_ABI_DT_RELRGLIBC_2.14GLIBC_2.4GLIBC_2.3.4GLIBC_2.2.5GLIBC_2.38GLIBC_2.29GLIBC_2.27 K`B U gii rti |ui Aui x= = = =  = = = = = = = = != = = &= )=  > *> +> (> B > D(> F0> 8> @> SH> VP> X> X`> %h> \p> ]x> ^> _> e> f> "> k> p> r> > > > > > '> > > ? ? ? ?  ? (? 0? 8? @? H? P? X? `? h? p? x?  ? ? $? ? ? ? ? ? ? ? ? #? &? ? 6 6 6 6 7  7  7  7  7  (7 07 87 @7 H7 P7 X7 `7 h7 p7 x7 7 7 7 7 7 7 7 7  7 "7 7 #7 $7 %7 '7 (7  8 ,8 -8 .8 / 8 0(8 108 288 3@8 4H8 5P8 6X8 7`8 8h8 9p8 :x8 ;8 <8 =8 >8 ?8 @8 A8 C8 E8 G8 H8 I8 J8 K8 L8 M8 N9 O9 P9 Q9 R 9 T(9 U09 W89 @9 YH9 ZP9 [X9 \`9 `h9 ap9 bx9 c9 d9 g9 h9 i9 j9 l9 m9 n9 o9 q9 s9 )9 t9 u9 v9 w: x: y: z: { : |(: }0: ~8: @: H: P: X: `: h: p: x: : : : : : : : : : : : : : : : : ; ; ; ;  ; (; 0; 8; @; H; P; X; `; h; p; x; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; < < < <  < (< 0< 8< @< H< P< X< `< h< p< x< < < < < < < < < < < < < < &<  < < = = = =  = (= 0= 8=  @= H= P= X= `= h= p=  UUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUU @ p9119  (1c3 00g#ffݻwwwwwwwwݝ2$C`( $ᡠPyObject_GetBuffer: view==NULL argument is obsoleteView.MemoryView.array.__getbuffer__View.MemoryView.memoryview.__getbuffer__View.MemoryView.array.get_memviewView.MemoryView.memoryview.setitem_indexedView.MemoryView.array.__getitem__%.200s() takes %.8s %zd positional argument%.1s (%zd given)View.MemoryView.array.__reduce_cython__View.MemoryView.memoryview.__reduce_cython__View.MemoryView._memoryviewslice.__reduce_cython__View.MemoryView.array.__setstate_cython__View.MemoryView.memoryview.__setstate_cython__View.MemoryView._memoryviewslice.__setstate_cython__View.MemoryView.Enum.__reduce_cython__View.MemoryView.memoryview.get_item_pointerinteger division or modulo by zerovalue too large to perform divisionView.MemoryView.pybuffer_indexView.MemoryView.memoryview.is_sliceView.MemoryView.memoryview.convert_item_to_objectView.MemoryView._memoryviewslice.convert_item_to_object while calling a Python objectNULL result without error in PyObject_CallView.MemoryView.memoryview.assign_item_from_objectView.MemoryView._memoryviewslice.assign_item_from_objectView.MemoryView.memoryview.__repr__View.MemoryView.memoryview.__str__Acquisition count is %d (line %d)View.MemoryView.memoryview_fromslice'NoneType' object is not subscriptableView.MemoryView.__pyx_unpickle_Enum__set_stateView.MemoryView.Enum.__setstate_cython__View.MemoryView.__pyx_unpickle_EnumView.MemoryView.array.__getattr__View.MemoryView.memoryview_cwrapper'NoneType' object is not iterableView.MemoryView.transpose_memsliceCannot convert %.200s to %.200sView.MemoryView.slice_memviewslicelocal variable '%s' referenced before assignmentView.MemoryView.memoryview.__getitem__too many values to unpack (expected %zd)need more than %zd value%.1s to unpackView.MemoryView.get_slice_from_memviewView.MemoryView.memoryview.is_c_contigView.MemoryView.memoryview.is_f_contigView.MemoryView.memoryview_copy_from_sliceView.MemoryView.memoryview.copyView.MemoryView.memoryview.copy_fortranView.MemoryView.assert_direct_dimensionsView.MemoryView.memoryview.setitem_slice_assign_scalarView.MemoryView._err_no_memoryView.MemoryView.copy_data_to_tempView.MemoryView.memoryview_copy_contentsView.MemoryView.memoryview.setitem_slice_assignment_ARRAY_API is not PyCapsule objectmodule compiled against ABI version 0x%x but this version of numpy is 0x%xmodule was compiled against NumPy C-API version 0x%x (NumPy 1.23) but the running NumPy has C-API version 0x%x. Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem.FATAL: module compiled as unknown endianFATAL: module compiled as little endian, but detected different endianness at runtimenumpy.random._generator.Generator.betanumpy.random._generator.Generator.exponentialnumpy.random._generator.Generator.__init__numpy.random._generator.Generator.__repr__numpy.random._generator.Generator.__str__numpy.random._generator.Generator.__setstate__numpy.random._generator.Generator.__reduce__numpy.random._generator.Generator.spawnnumpy.random._generator.Generator.randomnumpy.random._generator.Generator.integersnumpy.random._generator.Generator.standard_exponential'%.200s' object is unsliceablenumpy.random._generator.Generator.bytesnumpy.random._generator.Generator.choicenumpy.random._generator.Generator.uniform'%.200s' object does not support slice %.10snumpy.random._generator.Generator.normalnumpy.random._generator.Generator.gammanumpy.random._generator.Generator.fnumpy.random._generator.Generator.noncentral_fnumpy.random._generator.Generator.chisquarenumpy.random._generator.Generator.noncentral_chisquarenumpy.random._generator.Generator.standard_cauchynumpy.random._generator.Generator.standard_tnumpy.random._generator.Generator.vonmisesnumpy.random._generator.Generator.paretonumpy.random._generator.Generator.weibullnumpy.random._generator.Generator.powernumpy.random._generator.Generator.laplacenumpy.random._generator.Generator.gumbelnumpy.random._generator.Generator.logisticnumpy.random._generator.Generator.lognormalnumpy.random._generator.Generator.rayleighnumpy.random._generator.Generator.waldnumpy.random._generator.Generator.standard_normalnumpy.random._generator.Generator.standard_gammanumpy.random._generator.Generator.triangularnumpy.random._generator.Generator.poissonnumpy.random._generator.Generator.zipfnumpy.random._generator.Generator.geometricnumpy.random._generator.Generator.logseriesnumpy.random._generator.Generator.negative_binomialnumpy.random._generator.Generator.hypergeometricnumpy.random._generator.Generator.binomialnumpy.random._generator.Generator.multivariate_normalnumpy.random._generator.Generator.multinomialnumpy.random._generator.Generator.dirichletnumpy.random._generator.Generator.permutednumpy.random._generator.Generator.shufflenumpy.random._generator.Generator.multivariate_hypergeometricView.MemoryView.array.memview.__get__View.MemoryView.memoryview.base.__get__View.MemoryView.memoryview.ndim.__get__View.MemoryView.array.__setitem__Subscript deletion not supported by %.200snumpy.random._generator.Generator.permutationnumpy.random._generator.default_rngnumpy.random._generator._check_bit_generatorView.MemoryView.array.__cinit__Argument '%.200s' has incorrect type (expected %.200s, got %.200s)Argument '%.200s' must not be Noneobject of type 'NoneType' has no len()expected bytes, NoneType foundView.MemoryView._allocate_bufferView.MemoryView.memoryview.__cinit__View.MemoryView.memoryview.itemsize.__get__View.MemoryView.memoryview.size.__get__View.MemoryView.memoryview.__setitem__View.MemoryView.memoryview_copyView.MemoryView.memoryview.T.__get__View.MemoryView.memoryview.shape.__get__View.MemoryView.memoryview.strides.__get__View.MemoryView.memoryview.suboffsets.__get__View.MemoryView.memoryview.nbytes.__get__numpy.random._bounded_integersPyObject *(PyObject *, PyObject *, PyObject *, int, int, bitgen_t *, PyObject *)int (double, PyObject *, __pyx_t_5numpy_6random_7_common_constraint_type)int (PyArrayObject *, PyObject *, __pyx_t_5numpy_6random_7_common_constraint_type)PyObject *(void *, bitgen_t *, PyObject *, PyObject *, PyObject *)PyObject *(PyObject *, PyArrayObject *)PyObject *(void *, void *, PyObject *, PyObject *, int, PyObject *, PyObject *, __pyx_t_5numpy_6random_7_common_constraint_type, PyObject *, PyObject *, __pyx_t_5numpy_6random_7_common_constraint_type, PyObject *, PyObject *, __pyx_t_5numpy_6random_7_common_constraint_type, PyObject *)PyObject *(void *, void *, PyObject *, PyObject *, int, int, PyObject *, PyObject *, __pyx_t_5numpy_6random_7_common_constraint_type, PyObject *, PyObject *, __pyx_t_5numpy_6random_7_common_constraint_type, PyObject *, PyObject *, __pyx_t_5numpy_6random_7_common_constraint_type)PyObject *(void *, bitgen_t *, PyObject *, PyObject *, PyObject *, PyObject *, __pyx_t_5numpy_6random_7_common_constraint_type, PyObject *)PyObject *(void *, void *, PyObject *, PyObject *, PyArrayObject *, PyObject *, __pyx_t_5numpy_6random_7_common_constraint_type, PyArrayObject *, PyObject *, __pyx_t_5numpy_6random_7_common_constraint_type, PyArrayObject *, PyObject *, __pyx_t_5numpy_6random_7_common_constraint_type)__name__ must be set to a string object__qualname__ must be set to a string objectfunction's dictionary may not be deletedsetting function's dictionary to a non-dict__annotations__ must be set to a dict objectInterpreter change detected - this module can only be loaded into one interpreter per process.__defaults__ must be set to a tuple objectchanges to cyfunction.__defaults__ will not currently affect the values used in function calls__kwdefaults__ must be set to a dict objectchanges to cyfunction.__kwdefaults__ will not currently affect the values used in function callsShared Cython type %.200s is not a type objectShared Cython type %.200s has the wrong size, try recompiling%s() got multiple values for keyword argument '%U'%.200s() keywords must be stringsbase class '%.200s' is not a heap typeextension type '%.200s' has no __dict__ slot, but base type '%.200s' has: either add 'cdef dict __dict__' to the extension type or add '__slots__ = [...]' to the base typeinvalid vtable found for imported typemultiple bases have vtable conflict: '%.200s' and '%.200s'%.200s.%.200s is not a type object%.200s.%.200s size changed, may indicate binary incompatibility. Expected %zd from C header, got %zd from PyObjectunbound method %.200S() needs an argumentUnable to initialize pickling for %.200s%s() got an unexpected keyword argument '%U'instance exception may not have a separate valuecalling %R should have returned an instance of BaseException, not %Rraise: exception class must be a subclass of BaseExceptionexception causes must derive from BaseExceptionjoin() result is too long for a Python stringcannot fit '%.200s' into an index-sized integer'%.200s' object is not subscriptableModule '_generator' has already been imported. Re-initialisation is not supported.compile time Python version %d.%d of module '%.100s' %s runtime version %d.%d Gets the bit generator instance used by the generator Returns ------- bit_generator : BitGenerator The bit generator instance used by the generator _cython_3_1_3._common_types_metatype_cython_3_1_3.cython_function_or_methodnumpy.random._generator._memoryviewsliceInternal class for passing memoryview slices to Pythonnumpy.random._generator.memoryviewnumpy.random._generator.Generator Generator(bit_generator) Container for the BitGenerators. `Generator` exposes a number of methods for generating random numbers drawn from a variety of probability distributions. In addition to the distribution-specific arguments, each method takes a keyword argument `size` that defaults to ``None``. If `size` is ``None``, then a single value is generated and returned. If `size` is an integer, then a 1-D array filled with generated values is returned. If `size` is a tuple, then an array with that shape is filled and returned. The function :func:`numpy.random.default_rng` will instantiate a `Generator` with numpy's default `BitGenerator`. **No Compatibility Guarantee** `Generator` does not provide a version compatibility guarantee. In particular, as better algorithms evolve the bit stream may change. Parameters ---------- bit_generator : BitGenerator BitGenerator to use as the core generator. Notes ----- The Python stdlib module :external+python:mod:`random` contains pseudo-random number generator with a number of methods that are similar to the ones available in `Generator`. It uses Mersenne Twister, and this bit generator can be accessed using `MT19937`. `Generator`, besides being NumPy-aware, has the advantage that it provides a much larger number of probability distributions to choose from. Examples -------- >>> from numpy.random import Generator, PCG64 >>> rng = Generator(PCG64()) >>> rng.standard_normal() -0.203 # random See Also -------- default_rng : Recommended constructor for `Generator`. Unexpected format string character: '%c'Buffer dtype mismatch, expected %s%s%s but got %sBuffer dtype mismatch, expected '%s' but got %s in '%s.%s'Expected a dimension of size %zu, got %zuExpected %d dimensions, got %dPython does not define a standard format string size for long double ('g')..Buffer dtype mismatch; next field is at offset %zd but %zd expectedBig-endian buffer not supported on little-endian compilerBuffer acquisition: Expected '{' after 'T'Cannot handle repeated arrays in format stringDoes not understand character buffer dtype format string ('%c')Expected a dimension of size %zu, got %dExpected a comma in format string, got '%c'Expected %d dimension(s), got %dUnexpected end of format string, expected ')'memviewslice is already initialized!Buffer has wrong number of dimensions (expected %d, got %d)Item size of buffer (%zu byte%s) does not match size of '%s' (%zu byte%s)Buffer exposes suboffsets but no stridesC-contiguous buffer is not indirect in dimension %dBuffer and memoryview are not contiguous in the same dimension.Buffer is not indirectly contiguous in dimension %d.Buffer is not indirectly accessible in dimension %d.Buffer not compatible with direct access in dimension %d.Cannot copy memoryview slice with indirect dimensions (axis %d)View.MemoryView.array_cwrapper%.200s does not export expected C variable %.200sC variable %.200s.%.200s has wrong signature (expected %.500s, got %.500s)%.200s does not export expected C function %.200sC function %.200s.%.200s has wrong signature (expected %.500s, got %.500s)__int__ returned non-int (type %.200s). The ability to return an instance of a strict subclass of int is deprecated, and may be removed in a future version of Python.__int__ returned non-int (type %.200s)value too large to convert to intcan't convert negative value to size_texactly__reduce_cython____setstate_cython__View.MemoryView.Enum.__init__name '%U' is not definedExpected %s, got %.200s'NoneType' is not iterableView.MemoryView._err_dimtuple__pyx_unpickle_Enuman integer is requiredView.MemoryView._unellipsifyView.MemoryView._errView.MemoryView.memview_sliceMissing type objectmemviewsliceobjis_c_contigis_f_contigcopycopy_fortranView.MemoryView._err_extentsnumpy._core._multiarray_umathnumpy.core._multiarray_umath_ARRAY_API_ARRAY_API is NULL pointernumpy/__init__.cython-30.pxdnumpy.import_arrayat leastat mostbetanumpy/random/_generator.pyxBitGenerator__getstate____setstate____reduce__spawnrandomintegersstandard_exponentialchoiceuniformassignmentuint64_tnoncentral_fnoncentral_chisquarestandard_cauchystandard_tvonmisesparetoweibullpowerlaplacegumbellogisticlognormalrayleighwaldstandard_normalstandard_gammatriangularpoissonzipflogseriesnegative_binomialnumpy.PyArray_MultiIterNew2numpy.PyArray_MultiIterNew3multivariate_normalvhumultinomialdirichletpermutedoutshufflemultivariate_hypergeometricpermutationdefault_rng__cinit__format_rand_uint64_rand_uint32_rand_uint16_rand_uint8_rand_bool_rand_int64_rand_int32_rand_int16_rand_int8numpy.random._commoncheck_constraintcheck_array_constraintdouble (double *, npy_intp)kahan_sumdouble_fillfloat_fillvalidate_output_shapecontdisccont_fcont_broadcast_3discrete_broadcast_iii__loader__loader__file__origin__package__parent__path__submodule_search_locations4294967296builtinscomplexnumpyflatiterbroadcastndarraygenericnumberunsignedintegerinexactcomplexfloatingflexiblecharacterufuncnumpy.random.bit_generatorSeedSequenceSeedlessSequence_cython_3_1_3takes no arguments%.200s() %s (%zd given)takes exactly one argumentBad call flags for CyFunctiontakes no keyword arguments%.200s() %skeywords must be stringscython_runtime__builtins__does not match__debug__doubleLEGACY_POISSON_LAM_MAXMAXSIZEView.MemoryViewinit numpy.random._generator__module__func_doc__doc__func_name__name____qualname__func_dict__dict__func_globals__globals__func_closure__closure__func_code__code__func_defaults__defaults____kwdefaults____annotations___is_coroutine__dictoffset____vectorcalloffset____weaklistoffset__Tbasestridessuboffsetsndimitemsizenbytesmemview__getattr___bit_generatornumpy.random._generator.Enumnumpy.random._generator.array'bool''char''signed char''unsigned char''short''unsigned short''int''unsigned int''long''unsigned long''long long''unsigned long long''complex float''float''complex double''double''complex long double''long double'a structPython objecta pointera stringendunparsable format stringneeds an argumentbuffer dtypeBuffer not C contiguous.__pyx_capi__cannot import name %S00010203040506070809101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899000102030405060710111213141516172021222324252627303132333435363740414243444546475051525354555657606162636465666770717273747576770123456789abcdef0123456789ABCDEFE-t|" m U < &   t ^ H ?333333?C$@eA??:0yE>qh??@@0C@~)@ lѿ3 ; @UUUUUU?"@m{?̶e*@-DT! @h㈵>.A-DT! -DT!@5gG@dg?= ףp=@n?[ m?h|?5?333333 @r?$~?B>٬ @r鷯?Q?Q?9v?(\@ffffff@.@4@x&?@?UUUUUU?a@X@`@|@@MA>@v@UUUUUU?dg?3?r?q?0@IIIIIؗ—Ǫp[F0IQ1QQPPP/؈s\ zipf(a, size=None) Draw samples from a Zipf distribution. Samples are drawn from a Zipf distribution with specified parameter `a` > 1. The Zipf distribution (also known as the zeta distribution) is a discrete probability distribution that satisfies Zipf's law: the frequency of an item is inversely proportional to its rank in a frequency table. Parameters ---------- a : float or array_like of floats Distribution parameter. Must be greater than 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``a`` is a scalar. Otherwise, ``np.array(a).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Zipf distribution. See Also -------- scipy.stats.zipf : probability density function, distribution, or cumulative density function, etc. Notes ----- The probability mass function (PMF) for the Zipf distribution is .. math:: p(k) = \frac{k^{-a}}{\zeta(a)}, for integers :math:`k \geq 1`, where :math:`\zeta` is the Riemann Zeta function. It is named for the American linguist George Kingsley Zipf, who noted that the frequency of any word in a sample of a language is inversely proportional to its rank in the frequency table. References ---------- .. [1] Zipf, G. K., "Selected Studies of the Principle of Relative Frequency in Language," Cambridge, MA: Harvard Univ. Press, 1932. Examples -------- Draw samples from the distribution: >>> a = 4.0 >>> n = 20000 >>> rng = np.random.default_rng() >>> s = rng.zipf(a, size=n) Display the histogram of the samples, along with the expected histogram based on the probability density function: >>> import matplotlib.pyplot as plt >>> from scipy.special import zeta # doctest: +SKIP `bincount` provides a fast histogram for small integers. >>> count = np.bincount(s) >>> k = np.arange(1, s.max() + 1) >>> plt.bar(k, count[1:], alpha=0.5, label='sample count') >>> plt.plot(k, n*(k**-a)/zeta(a), 'k.-', alpha=0.5, ... label='expected count') # doctest: +SKIP >>> plt.semilogy() >>> plt.grid(alpha=0.4) >>> plt.legend() >>> plt.title(f'Zipf sample, a={a}, size={n}') >>> plt.show() zipfzigzerosyou are shuffling a 'x_ptrxwriteable weibull(a, size=None) Draw samples from a Weibull distribution. Draw samples from a 1-parameter Weibull distribution with the given shape parameter `a`. .. math:: X = (-ln(U))^{1/a} Here, U is drawn from the uniform distribution over (0,1]. The more common 2-parameter Weibull, including a scale parameter :math:`\lambda` is just :math:`X = \lambda(-ln(U))^{1/a}`. Parameters ---------- a : float or array_like of floats Shape parameter of the distribution. Must be nonnegative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``a`` is a scalar. Otherwise, ``np.array(a).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Weibull distribution. See Also -------- scipy.stats.weibull_max scipy.stats.weibull_min scipy.stats.genextreme gumbel Notes ----- The Weibull (or Type III asymptotic extreme value distribution for smallest values, SEV Type III, or Rosin-Rammler distribution) is one of a class of Generalized Extreme Value (GEV) distributions used in modeling extreme value problems. This class includes the Gumbel and Frechet distributions. The probability density for the Weibull distribution is .. math:: p(x) = \frac{a} {\lambda}(\frac{x}{\lambda})^{a-1}e^{-(x/\lambda)^a}, where :math:`a` is the shape and :math:`\lambda` the scale. The function has its peak (the mode) at :math:`\lambda(\frac{a-1}{a})^{1/a}`. When ``a = 1``, the Weibull distribution reduces to the exponential distribution. References ---------- .. [1] Waloddi Weibull, Royal Technical University, Stockholm, 1939 "A Statistical Theory Of The Strength Of Materials", Ingeniorsvetenskapsakademiens Handlingar Nr 151, 1939, Generalstabens Litografiska Anstalts Forlag, Stockholm. .. [2] Waloddi Weibull, "A Statistical Distribution Function of Wide Applicability", Journal Of Applied Mechanics ASME Paper 1951. .. [3] Wikipedia, "Weibull distribution", https://en.wikipedia.org/wiki/Weibull_distribution Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> a = 5. # shape >>> s = rng.weibull(a, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> def weibull(x, n, a): ... return (a / n) * (x / n)**(a - 1) * np.exp(-(x / n)**a) >>> count, bins, _ = plt.hist(rng.weibull(5., 1000)) >>> x = np.linspace(0, 2, 1000) >>> bin_spacing = np.mean(np.diff(bins)) >>> plt.plot(x, weibull(x, 1., 5.) * bin_spacing * s.size, label='Weibull PDF') >>> plt.legend() >>> plt.show() weibullwarningswarn wald(mean, scale, size=None) Draw samples from a Wald, or inverse Gaussian, distribution. As the scale approaches infinity, the distribution becomes more like a Gaussian. Some references claim that the Wald is an inverse Gaussian with mean equal to 1, but this is by no means universal. The inverse Gaussian distribution was first studied in relationship to Brownian motion. In 1956 M.C.K. Tweedie used the name inverse Gaussian because there is an inverse relationship between the time to cover a unit distance and distance covered in unit time. Parameters ---------- mean : float or array_like of floats Distribution mean, must be > 0. scale : float or array_like of floats Scale parameter, must be > 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``mean`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(mean, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Wald distribution. Notes ----- The probability density function for the Wald distribution is .. math:: P(x;mean,scale) = \sqrt{\frac{scale}{2\pi x^3}}e^ \frac{-scale(x-mean)^2}{2\cdotp mean^2x} As noted above the inverse Gaussian distribution first arise from attempts to model Brownian motion. It is also a competitor to the Weibull for use in reliability modeling and modeling stock returns and interest rate processes. References ---------- .. [1] Brighton Webs Ltd., Wald Distribution, https://web.archive.org/web/20090423014010/http://www.brighton-webs.co.uk:80/distributions/wald.asp .. [2] Chhikara, Raj S., and Folks, J. Leroy, "The Inverse Gaussian Distribution: Theory : Methodology, and Applications", CRC Press, 1988. .. [3] Wikipedia, "Inverse Gaussian distribution" https://en.wikipedia.org/wiki/Inverse_Gaussian_distribution Examples -------- Draw values from the distribution and plot the histogram: >>> import matplotlib.pyplot as plt >>> rng = np.random.default_rng() >>> h = plt.hist(rng.wald(3, 2, 100000), bins=200, density=True) >>> plt.show() wald vonmises(mu, kappa, size=None) Draw samples from a von Mises distribution. Samples are drawn from a von Mises distribution with specified mode (mu) and concentration (kappa), on the interval [-pi, pi]. The von Mises distribution (also known as the circular normal distribution) is a continuous probability distribution on the unit circle. It may be thought of as the circular analogue of the normal distribution. Parameters ---------- mu : float or array_like of floats Mode ("center") of the distribution. kappa : float or array_like of floats Concentration of the distribution, has to be >=0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``mu`` and ``kappa`` are both scalars. Otherwise, ``np.broadcast(mu, kappa).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized von Mises distribution. See Also -------- scipy.stats.vonmises : probability density function, distribution, or cumulative density function, etc. Notes ----- The probability density for the von Mises distribution is .. math:: p(x) = \frac{e^{\kappa cos(x-\mu)}}{2\pi I_0(\kappa)}, where :math:`\mu` is the mode and :math:`\kappa` the concentration, and :math:`I_0(\kappa)` is the modified Bessel function of order 0. The von Mises is named for Richard Edler von Mises, who was born in Austria-Hungary, in what is now the Ukraine. He fled to the United States in 1939 and became a professor at Harvard. He worked in probability theory, aerodynamics, fluid mechanics, and philosophy of science. References ---------- .. [1] Abramowitz, M. and Stegun, I. A. (Eds.). "Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables, 9th printing," New York: Dover, 1972. .. [2] von Mises, R., "Mathematical Theory of Probability and Statistics", New York: Academic Press, 1964. Examples -------- Draw samples from the distribution: >>> mu, kappa = 0.0, 4.0 # mean and concentration >>> rng = np.random.default_rng() >>> s = rng.vonmises(mu, kappa, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> from scipy.special import i0 # doctest: +SKIP >>> plt.hist(s, 50, density=True) >>> x = np.linspace(-np.pi, np.pi, num=51) >>> y = np.exp(kappa*np.cos(x-mu))/(2*np.pi*i0(kappa)) # doctest: +SKIP >>> plt.plot(x, y, linewidth=2, color='r') # doctest: +SKIP >>> plt.show() vonmisesvhvariates_ptrvariatesval_dataval_arrvalvupdateunpackunique_indicesuniqueuniform_samples uniform(low=0.0, high=1.0, size=None) Draw samples from a uniform distribution. Samples are uniformly distributed over the half-open interval ``[low, high)`` (includes low, but excludes high). In other words, any value within the given interval is equally likely to be drawn by `uniform`. Parameters ---------- low : float or array_like of floats, optional Lower boundary of the output interval. All values generated will be greater than or equal to low. The default value is 0. high : float or array_like of floats Upper boundary of the output interval. All values generated will be less than high. The high limit may be included in the returned array of floats due to floating-point rounding in the equation ``low + (high-low) * random_sample()``. high - low must be non-negative. The default value is 1.0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``low`` and ``high`` are both scalars. Otherwise, ``np.broadcast(low, high).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized uniform distribution. See Also -------- integers : Discrete uniform distribution, yielding integers. random : Floats uniformly distributed over ``[0, 1)``. Notes ----- The probability density function of the uniform distribution is .. math:: p(x) = \frac{1}{b - a} anywhere within the interval ``[a, b)``, and zero elsewhere. When ``high`` == ``low``, values of ``low`` will be returned. Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> s = rng.uniform(-1,0,1000) All values are within the given interval: >>> np.all(s >= -1) True >>> np.all(s < 0) True Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, _ = plt.hist(s, 15, density=True) >>> plt.plot(bins, np.ones_like(bins), linewidth=2, color='r') >>> plt.show() uniformunable to allocate shape and strides.unable to allocate array data.uint8uint64uint32uint16>> import matplotlib.pyplot as plt >>> rng = np.random.default_rng() >>> h = plt.hist(rng.triangular(-3, 0, 8, 100000), bins=200, ... density=True) >>> plt.show() triangulartotsizetotaltoltobytesto_shuffle__test__temp_arrtemptaketszswapaxessvdsum(pvalssum(colors) must not exceed the maximum value of a 64 bit signed integer (%d)sumsubtractstructstridesstridestopstepstatusstatestart standard_t(df, size=None) Draw samples from a standard Student's t distribution with `df` degrees of freedom. A special case of the hyperbolic distribution. As `df` gets large, the result resembles that of the standard normal distribution (`standard_normal`). Parameters ---------- df : float or array_like of floats Degrees of freedom, must be > 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``df`` is a scalar. Otherwise, ``np.array(df).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized standard Student's t distribution. Notes ----- The probability density function for the t distribution is .. math:: P(x, df) = \frac{\Gamma(\frac{df+1}{2})}{\sqrt{\pi df} \Gamma(\frac{df}{2})}\Bigl( 1+\frac{x^2}{df} \Bigr)^{-(df+1)/2} The t test is based on an assumption that the data come from a Normal distribution. The t test provides a way to test whether the sample mean (that is the mean calculated from the data) is a good estimate of the true mean. The derivation of the t-distribution was first published in 1908 by William Gosset while working for the Guinness Brewery in Dublin. Due to proprietary issues, he had to publish under a pseudonym, and so he used the name Student. References ---------- .. [1] Dalgaard, Peter, "Introductory Statistics With R", Springer, 2002. .. [2] Wikipedia, "Student's t-distribution" https://en.wikipedia.org/wiki/Student's_t-distribution Examples -------- From Dalgaard page 83 [1]_, suppose the daily energy intake for 11 women in kilojoules (kJ) is: >>> intake = np.array([5260., 5470, 5640, 6180, 6390, 6515, 6805, 7515, \ ... 7515, 8230, 8770]) Does their energy intake deviate systematically from the recommended value of 7725 kJ? Our null hypothesis will be the absence of deviation, and the alternate hypothesis will be the presence of an effect that could be either positive or negative, hence making our test 2-tailed. Because we are estimating the mean and we have N=11 values in our sample, we have N-1=10 degrees of freedom. We set our significance level to 95% and compute the t statistic using the empirical mean and empirical standard deviation of our intake. We use a ddof of 1 to base the computation of our empirical standard deviation on an unbiased estimate of the variance (note: the final estimate is not unbiased due to the concave nature of the square root). >>> np.mean(intake) 6753.636363636364 >>> intake.std(ddof=1) 1142.1232221373727 >>> t = (np.mean(intake)-7725)/(intake.std(ddof=1)/np.sqrt(len(intake))) >>> t -2.8207540608310198 We draw 1000000 samples from Student's t distribution with the adequate degrees of freedom. >>> import matplotlib.pyplot as plt >>> rng = np.random.default_rng() >>> s = rng.standard_t(10, size=1000000) >>> h = plt.hist(s, bins=100, density=True) Does our t statistic land in one of the two critical regions found at both tails of the distribution? >>> np.sum(np.abs(t) < np.abs(s)) / float(len(s)) 0.018318 #random < 0.05, statistic is in critical region The probability value for this 2-tailed test is about 1.83%, which is lower than the 5% pre-determined significance threshold. Therefore, the probability of observing values as extreme as our intake conditionally on the null hypothesis being true is too low, and we reject the null hypothesis of no deviation. standard_t standard_normal(size=None, dtype=np.float64, out=None) Draw samples from a standard Normal distribution (mean=0, stdev=1). Parameters ---------- size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. dtype : dtype, optional Desired dtype of the result, only `float64` and `float32` are supported. Byteorder must be native. The default value is np.float64. out : ndarray, optional Alternative output array in which to place the result. If size is not None, it must have the same shape as the provided size and must match the type of the output values. Returns ------- out : float or ndarray A floating-point array of shape ``size`` of drawn samples, or a single sample if ``size`` was not specified. See Also -------- normal : Equivalent function with additional ``loc`` and ``scale`` arguments for setting the mean and standard deviation. Notes ----- For random samples from the normal distribution with mean ``mu`` and standard deviation ``sigma``, use one of:: mu + sigma * rng.standard_normal(size=...) rng.normal(mu, sigma, size=...) Examples -------- >>> rng = np.random.default_rng() >>> rng.standard_normal() 2.1923875335537315 # random >>> s = rng.standard_normal(8000) >>> s array([ 0.6888893 , 0.78096262, -0.89086505, ..., 0.49876311, # random -0.38672696, -0.4685006 ]) # random >>> s.shape (8000,) >>> s = rng.standard_normal(size=(3, 4, 2)) >>> s.shape (3, 4, 2) Two-by-four array of samples from the normal distribution with mean 3 and standard deviation 2.5: >>> 3 + 2.5 * rng.standard_normal(size=(2, 4)) array([[-4.49401501, 4.00950034, -1.81814867, 7.29718677], # random [ 0.39924804, 4.68456316, 4.99394529, 4.84057254]]) # random standard_normal standard_gamma(shape, size=None, dtype=np.float64, out=None) Draw samples from a standard Gamma distribution. Samples are drawn from a Gamma distribution with specified parameters, shape (sometimes designated "k") and scale=1. Parameters ---------- shape : float or array_like of floats Parameter, must be non-negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``shape`` is a scalar. Otherwise, ``np.array(shape).size`` samples are drawn. dtype : dtype, optional Desired dtype of the result, only `float64` and `float32` are supported. Byteorder must be native. The default value is np.float64. out : ndarray, optional Alternative output array in which to place the result. If size is not None, it must have the same shape as the provided size and must match the type of the output values. Returns ------- out : ndarray or scalar Drawn samples from the parameterized standard gamma distribution. See Also -------- scipy.stats.gamma : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Gamma distribution is .. math:: p(x) = x^{k-1}\frac{e^{-x/\theta}}{\theta^k\Gamma(k)}, where :math:`k` is the shape and :math:`\theta` the scale, and :math:`\Gamma` is the Gamma function. The Gamma distribution is often used to model the times to failure of electronic components, and arises naturally in processes for which the waiting times between Poisson distributed events are relevant. References ---------- .. [1] Weisstein, Eric W. "Gamma Distribution." From MathWorld--A Wolfram Web Resource. https://mathworld.wolfram.com/GammaDistribution.html .. [2] Wikipedia, "Gamma distribution", https://en.wikipedia.org/wiki/Gamma_distribution Examples -------- Draw samples from the distribution: >>> shape, scale = 2., 1. # mean and width >>> rng = np.random.default_rng() >>> s = rng.standard_gamma(shape, 1000000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> import scipy.special as sps # doctest: +SKIP >>> count, bins, _ = plt.hist(s, 50, density=True) >>> y = bins**(shape-1) * ((np.exp(-bins/scale))/ # doctest: +SKIP ... (sps.gamma(shape) * scale**shape)) >>> plt.plot(bins, y, linewidth=2, color='r') # doctest: +SKIP >>> plt.show() standard_gamma standard_exponential(size=None, dtype=np.float64, method='zig', out=None) Draw samples from the standard exponential distribution. `standard_exponential` is identical to the exponential distribution with a scale parameter of 1. Parameters ---------- size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. dtype : dtype, optional Desired dtype of the result, only `float64` and `float32` are supported. Byteorder must be native. The default value is np.float64. method : str, optional Either 'inv' or 'zig'. 'inv' uses the default inverse CDF method. 'zig' uses the much faster Ziggurat method of Marsaglia and Tsang. out : ndarray, optional Alternative output array in which to place the result. If size is not None, it must have the same shape as the provided size and must match the type of the output values. Returns ------- out : float or ndarray Drawn samples. Examples -------- Output a 3x8000 array: >>> rng = np.random.default_rng() >>> n = rng.standard_exponential((3, 8000)) standard_exponential standard_cauchy(size=None) Draw samples from a standard Cauchy distribution with mode = 0. Also known as the Lorentz distribution. Parameters ---------- size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. Returns ------- samples : ndarray or scalar The drawn samples. Notes ----- The probability density function for the full Cauchy distribution is .. math:: P(x; x_0, \gamma) = \frac{1}{\pi \gamma \bigl[ 1+ (\frac{x-x_0}{\gamma})^2 \bigr] } and the Standard Cauchy distribution just sets :math:`x_0=0` and :math:`\gamma=1` The Cauchy distribution arises in the solution to the driven harmonic oscillator problem, and also describes spectral line broadening. It also describes the distribution of values at which a line tilted at a random angle will cut the x axis. When studying hypothesis tests that assume normality, seeing how the tests perform on data from a Cauchy distribution is a good indicator of their sensitivity to a heavy-tailed distribution, since the Cauchy looks very much like a Gaussian distribution, but with heavier tails. References ---------- .. [1] NIST/SEMATECH e-Handbook of Statistical Methods, "Cauchy Distribution", https://www.itl.nist.gov/div898/handbook/eda/section3/eda3663.htm .. [2] Weisstein, Eric W. "Cauchy Distribution." From MathWorld--A Wolfram Web Resource. https://mathworld.wolfram.com/CauchyDistribution.html .. [3] Wikipedia, "Cauchy distribution" https://en.wikipedia.org/wiki/Cauchy_distribution Examples -------- Draw samples and plot the distribution: >>> import matplotlib.pyplot as plt >>> rng = np.random.default_rng() >>> s = rng.standard_cauchy(1000000) >>> s = s[(s>-25) & (s<25)] # truncate distribution so it plots well >>> plt.hist(s, bins=100) >>> plt.show() standard_cauchystacklevelsqrt__spec__ spawn(n_children) Create new independent child generators. See :ref:`seedsequence-spawn` for additional notes on spawning children. .. versionadded:: 1.25.0 Parameters ---------- n_children : int Returns ------- child_generators : list of Generators Raises ------ TypeError When the underlying SeedSequence does not implement spawning. See Also -------- random.BitGenerator.spawn, random.SeedSequence.spawn : Equivalent method on the bit generator and seed sequence. bit_generator : The bit generator instance used by the generator. Examples -------- Starting from a seeded default generator: >>> # High quality entropy created with: f"0x{secrets.randbits(128):x}" >>> entropy = 0x3034c61a9ae04ff8cb62ab8ec2c4b501 >>> rng = np.random.default_rng(entropy) Create two new generators for example for parallel execution: >>> child_rng1, child_rng2 = rng.spawn(2) Drawn numbers from each are independent but derived from the initial seeding entropy: >>> rng.uniform(), child_rng1.uniform(), child_rng2.uniform() (0.19029263503854454, 0.9475673279178444, 0.4702687338396767) It is safe to spawn additional children from the original ``rng`` or the children: >>> more_child_rngs = rng.spawn(20) >>> nested_spawn = child_rng1.spawn(20) spawnsortslicesslice_reprsize_isizesigmaside shuffle(x, axis=0) Modify an array or sequence in-place by shuffling its contents. The order of sub-arrays is changed but their contents remains the same. Parameters ---------- x : ndarray or MutableSequence The array, list or mutable sequence to be shuffled. axis : int, optional The axis which `x` is shuffled along. Default is 0. It is only supported on `ndarray` objects. Returns ------- None See Also -------- permuted permutation Notes ----- An important distinction between methods ``shuffle`` and ``permuted`` is how they both treat the ``axis`` parameter which can be found at :ref:`generator-handling-axis-parameter`. Examples -------- >>> rng = np.random.default_rng() >>> arr = np.arange(10) >>> arr array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9]) >>> rng.shuffle(arr) >>> arr array([2, 0, 7, 5, 1, 4, 8, 9, 3, 6]) # random >>> arr = np.arange(9).reshape((3, 3)) >>> arr array([[0, 1, 2], [3, 4, 5], [6, 7, 8]]) >>> rng.shuffle(arr) >>> arr array([[3, 4, 5], # random [6, 7, 8], [0, 1, 2]]) >>> arr = np.arange(9).reshape((3, 3)) >>> arr array([[0, 1, 2], [3, 4, 5], [6, 7, 8]]) >>> rng.shuffle(arr, axis=1) >>> arr array([[2, 0, 1], # random [5, 3, 4], [8, 6, 7]]) shuffleshape__setstate_cython____setstate__set_size__set_name__selfseedsearchsortedscalesafesrtolrngrightreversedreturn_indexretresultreshaperesreplaceregister__reduce_ex____reduce_cython____reduce__reduce rayleigh(scale=1.0, size=None) Draw samples from a Rayleigh distribution. The :math:`\chi` and Weibull distributions are generalizations of the Rayleigh. Parameters ---------- scale : float or array_like of floats, optional Scale, also equals the mode. Must be non-negative. Default is 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``scale`` is a scalar. Otherwise, ``np.array(scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Rayleigh distribution. Notes ----- The probability density function for the Rayleigh distribution is .. math:: P(x;scale) = \frac{x}{scale^2}e^{\frac{-x^2}{2 \cdotp scale^2}} The Rayleigh distribution would arise, for example, if the East and North components of the wind velocity had identical zero-mean Gaussian distributions. Then the wind speed would have a Rayleigh distribution. References ---------- .. [1] Brighton Webs Ltd., "Rayleigh Distribution," https://web.archive.org/web/20090514091424/http://brighton-webs.co.uk:80/distributions/rayleigh.asp .. [2] Wikipedia, "Rayleigh distribution" https://en.wikipedia.org/wiki/Rayleigh_distribution Examples -------- Draw values from the distribution and plot the histogram >>> from matplotlib.pyplot import hist >>> rng = np.random.default_rng() >>> values = hist(rng.rayleigh(3, 100000), bins=200, density=True) Wave heights tend to follow a Rayleigh distribution. If the mean wave height is 1 meter, what fraction of waves are likely to be larger than 3 meters? >>> meanvalue = 1 >>> modevalue = np.sqrt(2 / np.pi) * meanvalue >>> s = rng.rayleigh(modevalue, 1000000) The percentage of waves larger than 3 meters is: >>> 100.*sum(s>3)/1000000. 0.087300000000000003 # random rayleighravelrangerandoms_datarandoms random(size=None, dtype=np.float64, out=None) Return random floats in the half-open interval [0.0, 1.0). Results are from the "continuous uniform" distribution over the stated interval. To sample :math:`Unif[a, b), b > a` use `uniform` or multiply the output of `random` by ``(b - a)`` and add ``a``:: (b - a) * random() + a Parameters ---------- size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. dtype : dtype, optional Desired dtype of the result, only `float64` and `float32` are supported. Byteorder must be native. The default value is np.float64. out : ndarray, optional Alternative output array in which to place the result. If size is not None, it must have the same shape as the provided size and must match the type of the output values. Returns ------- out : float or ndarray of floats Array of random floats of shape `size` (unless ``size=None``, in which case a single float is returned). See Also -------- uniform : Draw samples from the parameterized uniform distribution. Examples -------- >>> rng = np.random.default_rng() >>> rng.random() 0.47108547995356098 # random >>> type(rng.random()) >>> rng.random((5,)) array([ 0.30220482, 0.86820401, 0.1654503 , 0.11659149, 0.54323428]) # random Three-by-two array of random numbers from [-5, 0): >>> 5 * rng.random((3, 2)) - 5 array([[-3.99149989, -0.52338984], # random [-2.99091858, -0.79479508], [-1.23204345, -1.75224494]]) randomraise__qualname____pyx_vtable____pyx_unpickle_Enum__pyx_type__pyx_state__pyx_checksumpvals must have at least 1 dimension and the last dimension of pvals must be greater than 0.pvalspsdprod power(a, size=None) Draws samples in [0, 1] from a power distribution with positive exponent a - 1. Also known as the power function distribution. Parameters ---------- a : float or array_like of floats Parameter of the distribution. Must be non-negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``a`` is a scalar. Otherwise, ``np.array(a).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized power distribution. Raises ------ ValueError If a <= 0. Notes ----- The probability density function is .. math:: P(x; a) = ax^{a-1}, 0 \le x \le 1, a>0. The power function distribution is just the inverse of the Pareto distribution. It may also be seen as a special case of the Beta distribution. It is used, for example, in modeling the over-reporting of insurance claims. References ---------- .. [1] Christian Kleiber, Samuel Kotz, "Statistical size distributions in economics and actuarial sciences", Wiley, 2003. .. [2] Heckert, N. A. and Filliben, James J. "NIST Handbook 148: Dataplot Reference Manual, Volume 2: Let Subcommands and Library Functions", National Institute of Standards and Technology Handbook Series, June 2003. https://www.itl.nist.gov/div898/software/dataplot/refman2/auxillar/powpdf.pdf Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> a = 5. # shape >>> samples = 1000 >>> s = rng.power(a, samples) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, _ = plt.hist(s, bins=30) >>> x = np.linspace(0, 1, 100) >>> y = a*x**(a-1.) >>> normed_y = samples*np.diff(bins)[0]*y >>> plt.plot(x, normed_y) >>> plt.show() Compare the power function distribution to the inverse of the Pareto. >>> from scipy import stats # doctest: +SKIP >>> rvs = rng.power(5, 1000000) >>> rvsp = rng.pareto(5, 1000000) >>> xx = np.linspace(0,1,100) >>> powpdf = stats.powerlaw.pdf(xx,5) # doctest: +SKIP >>> plt.figure() >>> plt.hist(rvs, bins=50, density=True) >>> plt.plot(xx,powpdf,'r-') # doctest: +SKIP >>> plt.title('power(5)') >>> plt.figure() >>> plt.hist(1./(1.+rvsp), bins=50, density=True) >>> plt.plot(xx,powpdf,'r-') # doctest: +SKIP >>> plt.title('inverse of 1 + Generator.pareto(5)') >>> plt.figure() >>> plt.hist(1./(1.+rvsp), bins=50, density=True) >>> plt.plot(xx,powpdf,'r-') # doctest: +SKIP >>> plt.title('inverse of stats.pareto(5)') powerpop_size_ipop_sizepop_poisson_lam_max poisson(lam=1.0, size=None) Draw samples from a Poisson distribution. The Poisson distribution is the limit of the binomial distribution for large N. Parameters ---------- lam : float or array_like of floats Expected number of events occurring in a fixed-time interval, must be >= 0. A sequence must be broadcastable over the requested size. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``lam`` is a scalar. Otherwise, ``np.array(lam).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Poisson distribution. Notes ----- The probability mass function (PMF) of Poisson distribution is .. math:: f(k; \lambda)=\frac{\lambda^k e^{-\lambda}}{k!} For events with an expected separation :math:`\lambda` the Poisson distribution :math:`f(k; \lambda)` describes the probability of :math:`k` events occurring within the observed interval :math:`\lambda`. Because the output is limited to the range of the C int64 type, a ValueError is raised when `lam` is within 10 sigma of the maximum representable value. References ---------- .. [1] Weisstein, Eric W. "Poisson Distribution." From MathWorld--A Wolfram Web Resource. https://mathworld.wolfram.com/PoissonDistribution.html .. [2] Wikipedia, "Poisson distribution", https://en.wikipedia.org/wiki/Poisson_distribution Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> lam, size = 5, 10000 >>> s = rng.poisson(lam=lam, size=size) Verify the mean and variance, which should be approximately ``lam``: >>> s.mean(), s.var() (4.9917 5.1088311) # may vary Display the histogram and probability mass function: >>> import matplotlib.pyplot as plt >>> from scipy import stats >>> x = np.arange(0, 21) >>> pmf = stats.poisson.pmf(x, mu=lam) >>> plt.hist(s, bins=x, density=True, width=0.5) >>> plt.stem(x, pmf, 'C1-') >>> plt.show() Draw each 100 values for lambda 100 and 500: >>> s = rng.poisson(lam=(100., 500.), size=(100, 2)) poissonpix_picklepicklepi permuted(x, axis=None, out=None) Randomly permute `x` along axis `axis`. Unlike `shuffle`, each slice along the given axis is shuffled independently of the others. Parameters ---------- x : array_like, at least one-dimensional Array to be shuffled. axis : int, optional Slices of `x` in this axis are shuffled. Each slice is shuffled independently of the others. If `axis` is None, the flattened array is shuffled. out : ndarray, optional If given, this is the destination of the shuffled array. If `out` is None, a shuffled copy of the array is returned. Returns ------- ndarray If `out` is None, a shuffled copy of `x` is returned. Otherwise, the shuffled array is stored in `out`, and `out` is returned See Also -------- shuffle permutation Notes ----- An important distinction between methods ``shuffle`` and ``permuted`` is how they both treat the ``axis`` parameter which can be found at :ref:`generator-handling-axis-parameter`. Examples -------- Create a `numpy.random.Generator` instance: >>> rng = np.random.default_rng() Create a test array: >>> x = np.arange(24).reshape(3, 8) >>> x array([[ 0, 1, 2, 3, 4, 5, 6, 7], [ 8, 9, 10, 11, 12, 13, 14, 15], [16, 17, 18, 19, 20, 21, 22, 23]]) Shuffle the rows of `x`: >>> y = rng.permuted(x, axis=1) >>> y array([[ 4, 3, 6, 7, 1, 2, 5, 0], # random [15, 10, 14, 9, 12, 11, 8, 13], [17, 16, 20, 21, 18, 22, 23, 19]]) `x` has not been modified: >>> x array([[ 0, 1, 2, 3, 4, 5, 6, 7], [ 8, 9, 10, 11, 12, 13, 14, 15], [16, 17, 18, 19, 20, 21, 22, 23]]) To shuffle the rows of `x` in-place, pass `x` as the `out` parameter: >>> y = rng.permuted(x, axis=1, out=x) >>> x array([[ 3, 0, 4, 7, 1, 6, 2, 5], # random [ 8, 14, 13, 9, 12, 11, 15, 10], [17, 18, 16, 22, 19, 23, 20, 21]]) Note that when the ``out`` parameter is given, the return value is ``out``: >>> y is x True permuted permutation(x, axis=0) Randomly permute a sequence, or return a permuted range. Parameters ---------- x : int or array_like If `x` is an integer, randomly permute ``np.arange(x)``. If `x` is an array, make a copy and shuffle the elements randomly. axis : int, optional The axis which `x` is shuffled along. Default is 0. Returns ------- out : ndarray Permuted sequence or array range. Examples -------- >>> rng = np.random.default_rng() >>> rng.permutation(10) array([1, 7, 4, 3, 0, 9, 2, 5, 8, 6]) # random >>> rng.permutation([1, 4, 9, 12, 15]) array([15, 1, 9, 4, 12]) # random >>> arr = np.arange(9).reshape((3, 3)) >>> rng.permutation(arr) array([[6, 7, 8], # random [0, 1, 2], [3, 4, 5]]) >>> rng.permutation("abc") Traceback (most recent call last): ... numpy.exceptions.AxisError: axis 0 is out of bounds for array of dimension 0 >>> arr = np.arange(9).reshape((3, 3)) >>> rng.permutation(arr, axis=1) array([[0, 2, 1], # random [3, 5, 4], [6, 8, 7]]) permutation_pcg64parr pareto(a, size=None) Draw samples from a Pareto II (AKA Lomax) distribution with specified shape. Parameters ---------- a : float or array_like of floats Shape of the distribution. Must be positive. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``a`` is a scalar. Otherwise, ``np.array(a).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the Pareto II distribution. See Also -------- scipy.stats.pareto : Pareto I distribution scipy.stats.lomax : Lomax (Pareto II) distribution scipy.stats.genpareto : Generalized Pareto distribution Notes ----- The probability density for the Pareto II distribution is .. math:: p(x) = \frac{a}{{x+1}^{a+1}} , x \ge 0 where :math:`a > 0` is the shape. The Pareto II distribution is a shifted and scaled version of the Pareto I distribution, which can be found in `scipy.stats.pareto`. References ---------- .. [1] Francis Hunt and Paul Johnson, On the Pareto Distribution of Sourceforge projects. .. [2] Pareto, V. (1896). Course of Political Economy. Lausanne. .. [3] Reiss, R.D., Thomas, M.(2001), Statistical Analysis of Extreme Values, Birkhauser Verlag, Basel, pp 23-30. .. [4] Wikipedia, "Pareto distribution", https://en.wikipedia.org/wiki/Pareto_distribution Examples -------- Draw samples from the distribution: >>> a = 3. >>> rng = np.random.default_rng() >>> s = rng.pareto(a, 10000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> x = np.linspace(0, 3, 50) >>> pdf = a / (x+1)**(a+1) >>> plt.hist(s, bins=x, density=True, label='histogram') >>> plt.plot(x, pdf, linewidth=2, color='r', label='pdf') >>> plt.xlim(x.min(), x.max()) >>> plt.legend() >>> plt.show() paretopackp_sump must be 1-dimensionalp_arrpout must have the same shape as xout must be a numpy arrayoutorightorderoperatoronsampleongoodonbadonomodeoleftoffsetsoffset' object which is not a subclass of 'Sequence'; `shuffle` is not guaranteed to behave correctly. E.g., non-numpy array/tensor objects with view semantics may contain duplicates after shuffling. object>objnumpy/random/_generator.pyxnumpy.random._generatornumpy.randomnumpy.linalgnumpy.lib.array_utilsnumpy._core.umath failed to importnumpy._core.multiarray failed to importnumpynum_variatesnum_colorsnsample > sum(colors)nsample must not exceed %dnsample must be nonnegative.nsample must be an integernsamplensampnpnormalize_axis_index normal(loc=0.0, scale=1.0, size=None) Draw random samples from a normal (Gaussian) distribution. The probability density function of the normal distribution, first derived by De Moivre and 200 years later by both Gauss and Laplace independently [2]_, is often called the bell curve because of its characteristic shape (see the example below). The normal distributions occurs often in nature. For example, it describes the commonly occurring distribution of samples influenced by a large number of tiny, random disturbances, each with its own unique distribution [2]_. Parameters ---------- loc : float or array_like of floats Mean ("centre") of the distribution. scale : float or array_like of floats Standard deviation (spread or "width") of the distribution. Must be non-negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``loc`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(loc, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized normal distribution. See Also -------- scipy.stats.norm : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Gaussian distribution is .. math:: p(x) = \frac{1}{\sqrt{ 2 \pi \sigma^2 }} e^{ - \frac{ (x - \mu)^2 } {2 \sigma^2} }, where :math:`\mu` is the mean and :math:`\sigma` the standard deviation. The square of the standard deviation, :math:`\sigma^2`, is called the variance. The function has its peak at the mean, and its "spread" increases with the standard deviation (the function reaches 0.607 times its maximum at :math:`x + \sigma` and :math:`x - \sigma` [2]_). This implies that :meth:`normal` is more likely to return samples lying close to the mean, rather than those far away. References ---------- .. [1] Wikipedia, "Normal distribution", https://en.wikipedia.org/wiki/Normal_distribution .. [2] P. R. Peebles Jr., "Central Limit Theorem" in "Probability, Random Variables and Random Signal Principles", 4th ed., 2001, pp. 51, 51, 125. Examples -------- Draw samples from the distribution: >>> mu, sigma = 0, 0.1 # mean and standard deviation >>> rng = np.random.default_rng() >>> s = rng.normal(mu, sigma, 1000) Verify the mean and the standard deviation: >>> abs(mu - np.mean(s)) 0.0 # may vary >>> abs(sigma - np.std(s, ddof=1)) 0.0 # may vary Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, _ = plt.hist(s, 30, density=True) >>> plt.plot(bins, 1/(sigma * np.sqrt(2 * np.pi)) * ... np.exp( - (bins - mu)**2 / (2 * sigma**2) ), ... linewidth=2, color='r') >>> plt.show() Two-by-four array of samples from the normal distribution with mean 3 and standard deviation 2.5: >>> rng = np.random.default_rng() >>> rng.normal(3, 2.5, size=(2, 4)) array([[-4.49401501, 4.00950034, -1.81814867, 7.29718677], # random [ 0.39924804, 4.68456316, 4.99394529, 4.84057254]]) # random normal noncentral_f(dfnum, dfden, nonc, size=None) Draw samples from the noncentral F distribution. Samples are drawn from an F distribution with specified parameters, `dfnum` (degrees of freedom in numerator) and `dfden` (degrees of freedom in denominator), where both parameters > 1. `nonc` is the non-centrality parameter. Parameters ---------- dfnum : float or array_like of floats Numerator degrees of freedom, must be > 0. dfden : float or array_like of floats Denominator degrees of freedom, must be > 0. nonc : float or array_like of floats Non-centrality parameter, the sum of the squares of the numerator means, must be >= 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``dfnum``, ``dfden``, and ``nonc`` are all scalars. Otherwise, ``np.broadcast(dfnum, dfden, nonc).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized noncentral Fisher distribution. Notes ----- When calculating the power of an experiment (power = probability of rejecting the null hypothesis when a specific alternative is true) the non-central F statistic becomes important. When the null hypothesis is true, the F statistic follows a central F distribution. When the null hypothesis is not true, then it follows a non-central F statistic. References ---------- .. [1] Weisstein, Eric W. "Noncentral F-Distribution." From MathWorld--A Wolfram Web Resource. https://mathworld.wolfram.com/NoncentralF-Distribution.html .. [2] Wikipedia, "Noncentral F-distribution", https://en.wikipedia.org/wiki/Noncentral_F-distribution Examples -------- In a study, testing for a specific alternative to the null hypothesis requires use of the Noncentral F distribution. We need to calculate the area in the tail of the distribution that exceeds the value of the F distribution for the null hypothesis. We'll plot the two probability distributions for comparison. >>> rng = np.random.default_rng() >>> dfnum = 3 # between group deg of freedom >>> dfden = 20 # within groups degrees of freedom >>> nonc = 3.0 >>> nc_vals = rng.noncentral_f(dfnum, dfden, nonc, 1000000) >>> NF = np.histogram(nc_vals, bins=50, density=True) >>> c_vals = rng.f(dfnum, dfden, 1000000) >>> F = np.histogram(c_vals, bins=50, density=True) >>> import matplotlib.pyplot as plt >>> plt.plot(F[1][1:], F[0]) >>> plt.plot(NF[1][1:], NF[0]) >>> plt.show() noncentral_f noncentral_chisquare(df, nonc, size=None) Draw samples from a noncentral chi-square distribution. The noncentral :math:`\chi^2` distribution is a generalization of the :math:`\chi^2` distribution. Parameters ---------- df : float or array_like of floats Degrees of freedom, must be > 0. nonc : float or array_like of floats Non-centrality, must be non-negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``df`` and ``nonc`` are both scalars. Otherwise, ``np.broadcast(df, nonc).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized noncentral chi-square distribution. Notes ----- The probability density function for the noncentral Chi-square distribution is .. math:: P(x;df,nonc) = \sum^{\infty}_{i=0} \frac{e^{-nonc/2}(nonc/2)^{i}}{i!} P_{Y_{df+2i}}(x), where :math:`Y_{q}` is the Chi-square with q degrees of freedom. References ---------- .. [1] Wikipedia, "Noncentral chi-squared distribution" https://en.wikipedia.org/wiki/Noncentral_chi-squared_distribution Examples -------- Draw values from the distribution and plot the histogram >>> rng = np.random.default_rng() >>> import matplotlib.pyplot as plt >>> values = plt.hist(rng.noncentral_chisquare(3, 20, 100000), ... bins=200, density=True) >>> plt.show() Draw values from a noncentral chisquare with very small noncentrality, and compare to a chisquare. >>> plt.figure() >>> values = plt.hist(rng.noncentral_chisquare(3, .0000001, 100000), ... bins=np.arange(0., 25, .1), density=True) >>> values2 = plt.hist(rng.chisquare(3, 100000), ... bins=np.arange(0., 25, .1), density=True) >>> plt.plot(values[1][0:-1], values[0]-values2[0], 'ob') >>> plt.show() Demonstrate how large values of non-centrality lead to a more symmetric distribution. >>> plt.figure() >>> values = plt.hist(rng.noncentral_chisquare(3, 20, 100000), ... bins=200, density=True) >>> plt.show() noncentral_chisquarenoncno default __reduce__ due to non-trivial __cinit__ningood + nbad < nsamplengoodnew__new__negative dimensions are not allowed negative_binomial(n, p, size=None) Draw samples from a negative binomial distribution. Samples are drawn from a negative binomial distribution with specified parameters, `n` successes and `p` probability of success where `n` is > 0 and `p` is in the interval (0, 1]. Parameters ---------- n : float or array_like of floats Parameter of the distribution, > 0. p : float or array_like of floats Parameter of the distribution. Must satisfy 0 < p <= 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``n`` and ``p`` are both scalars. Otherwise, ``np.broadcast(n, p).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized negative binomial distribution, where each sample is equal to N, the number of failures that occurred before a total of n successes was reached. Notes ----- The probability mass function of the negative binomial distribution is .. math:: P(N;n,p) = \frac{\Gamma(N+n)}{N!\Gamma(n)}p^{n}(1-p)^{N}, where :math:`n` is the number of successes, :math:`p` is the probability of success, :math:`N+n` is the number of trials, and :math:`\Gamma` is the gamma function. When :math:`n` is an integer, :math:`\frac{\Gamma(N+n)}{N!\Gamma(n)} = \binom{N+n-1}{N}`, which is the more common form of this term in the pmf. The negative binomial distribution gives the probability of N failures given n successes, with a success on the last trial. If one throws a die repeatedly until the third time a "1" appears, then the probability distribution of the number of non-"1"s that appear before the third "1" is a negative binomial distribution. Because this method internally calls ``Generator.poisson`` with an intermediate random value, a ValueError is raised when the choice of :math:`n` and :math:`p` would result in the mean + 10 sigma of the sampled intermediate distribution exceeding the max acceptable value of the ``Generator.poisson`` method. This happens when :math:`p` is too low (a lot of failures happen for every success) and :math:`n` is too big ( a lot of successes are allowed). Therefore, the :math:`n` and :math:`p` values must satisfy the constraint: .. math:: n\frac{1-p}{p}+10n\sqrt{n}\frac{1-p}{p}<2^{63}-1-10\sqrt{2^{63}-1}, Where the left side of the equation is the derived mean + 10 sigma of a sample from the gamma distribution internally used as the :math:`lam` parameter of a poisson sample, and the right side of the equation is the constraint for maximum value of :math:`lam` in ``Generator.poisson``. References ---------- .. [1] Weisstein, Eric W. "Negative Binomial Distribution." From MathWorld--A Wolfram Web Resource. https://mathworld.wolfram.com/NegativeBinomialDistribution.html .. [2] Wikipedia, "Negative binomial distribution", https://en.wikipedia.org/wiki/Negative_binomial_distribution Examples -------- Draw samples from the distribution: A real world example. A company drills wild-cat oil exploration wells, each with an estimated probability of success of 0.1. What is the probability of having one success for each successive well, that is what is the probability of a single success after drilling 5 wells, after 6 wells, etc.? >>> rng = np.random.default_rng() >>> s = rng.negative_binomial(1, 0.1, 100000) >>> for i in range(1, 11): # doctest: +SKIP ... probability = sum(s>> mean = [0, 0] >>> cov = [[1, 0], [0, 100]] # diagonal covariance Diagonal covariance means that points are oriented along x or y-axis: >>> import matplotlib.pyplot as plt >>> rng = np.random.default_rng() >>> x, y = rng.multivariate_normal(mean, cov, 5000).T >>> plt.plot(x, y, 'x') >>> plt.axis('equal') >>> plt.show() Note that the covariance matrix must be positive semidefinite (a.k.a. nonnegative-definite). Otherwise, the behavior of this method is undefined and backwards compatibility is not guaranteed. This function internally uses linear algebra routines, and thus results may not be identical (even up to precision) across architectures, OSes, or even builds. For example, this is likely if ``cov`` has multiple equal singular values and ``method`` is ``'svd'`` (default). In this case, ``method='cholesky'`` may be more robust. References ---------- .. [1] Papoulis, A., "Probability, Random Variables, and Stochastic Processes," 3rd ed., New York: McGraw-Hill, 1991. .. [2] Duda, R. O., Hart, P. E., and Stork, D. G., "Pattern Classification," 2nd ed., New York: Wiley, 2001. Examples -------- >>> mean = (1, 2) >>> cov = [[1, 0], [0, 1]] >>> rng = np.random.default_rng() >>> x = rng.multivariate_normal(mean, cov, (3, 3)) >>> x.shape (3, 3, 2) We can use a different method other than the default to factorize cov: >>> y = rng.multivariate_normal(mean, cov, (3, 3), method='cholesky') >>> y.shape (3, 3, 2) Here we generate 800 samples from the bivariate normal distribution with mean [0, 0] and covariance matrix [[6, -3], [-3, 3.5]]. The expected variances of the first and second components of the sample are 6 and 3.5, respectively, and the expected correlation coefficient is -3/sqrt(6*3.5) ≈ -0.65465. >>> cov = np.array([[6, -3], [-3, 3.5]]) >>> pts = rng.multivariate_normal([0, 0], cov, size=800) Check that the mean, covariance, and correlation coefficient of the sample are close to the expected values: >>> pts.mean(axis=0) array([ 0.0326911 , -0.01280782]) # may vary >>> np.cov(pts.T) array([[ 5.96202397, -2.85602287], [-2.85602287, 3.47613949]]) # may vary >>> np.corrcoef(pts.T)[0, 1] -0.6273591314603949 # may vary We can visualize this data with a scatter plot. The orientation of the point cloud illustrates the negative correlation of the components of this sample. >>> import matplotlib.pyplot as plt >>> plt.plot(pts[:, 0], pts[:, 1], '.', alpha=0.5) >>> plt.axis('equal') >>> plt.grid() >>> plt.show() multivariate_normal multivariate_hypergeometric(colors, nsample, size=None, method='marginals') Generate variates from a multivariate hypergeometric distribution. The multivariate hypergeometric distribution is a generalization of the hypergeometric distribution. Choose ``nsample`` items at random without replacement from a collection with ``N`` distinct types. ``N`` is the length of ``colors``, and the values in ``colors`` are the number of occurrences of that type in the collection. The total number of items in the collection is ``sum(colors)``. Each random variate generated by this function is a vector of length ``N`` holding the counts of the different types that occurred in the ``nsample`` items. The name ``colors`` comes from a common description of the distribution: it is the probability distribution of the number of marbles of each color selected without replacement from an urn containing marbles of different colors; ``colors[i]`` is the number of marbles in the urn with color ``i``. Parameters ---------- colors : sequence of integers The number of each type of item in the collection from which a sample is drawn. The values in ``colors`` must be nonnegative. To avoid loss of precision in the algorithm, ``sum(colors)`` must be less than ``10**9`` when `method` is "marginals". nsample : int The number of items selected. ``nsample`` must not be greater than ``sum(colors)``. size : int or tuple of ints, optional The number of variates to generate, either an integer or a tuple holding the shape of the array of variates. If the given size is, e.g., ``(k, m)``, then ``k * m`` variates are drawn, where one variate is a vector of length ``len(colors)``, and the return value has shape ``(k, m, len(colors))``. If `size` is an integer, the output has shape ``(size, len(colors))``. Default is None, in which case a single variate is returned as an array with shape ``(len(colors),)``. method : string, optional Specify the algorithm that is used to generate the variates. Must be 'count' or 'marginals' (the default). See the Notes for a description of the methods. Returns ------- variates : ndarray Array of variates drawn from the multivariate hypergeometric distribution. See Also -------- hypergeometric : Draw samples from the (univariate) hypergeometric distribution. Notes ----- The two methods do not return the same sequence of variates. The "count" algorithm is roughly equivalent to the following numpy code:: choices = np.repeat(np.arange(len(colors)), colors) selection = np.random.choice(choices, nsample, replace=False) variate = np.bincount(selection, minlength=len(colors)) The "count" algorithm uses a temporary array of integers with length ``sum(colors)``. The "marginals" algorithm generates a variate by using repeated calls to the univariate hypergeometric sampler. It is roughly equivalent to:: variate = np.zeros(len(colors), dtype=np.int64) # `remaining` is the cumulative sum of `colors` from the last # element to the first; e.g. if `colors` is [3, 1, 5], then # `remaining` is [9, 6, 5]. remaining = np.cumsum(colors[::-1])[::-1] for i in range(len(colors)-1): if nsample < 1: break variate[i] = hypergeometric(colors[i], remaining[i+1], nsample) nsample -= variate[i] variate[-1] = nsample The default method is "marginals". For some cases (e.g. when `colors` contains relatively small integers), the "count" method can be significantly faster than the "marginals" method. If performance of the algorithm is important, test the two methods with typical inputs to decide which works best. Examples -------- >>> colors = [16, 8, 4] >>> seed = 4861946401452 >>> gen = np.random.Generator(np.random.PCG64(seed)) >>> gen.multivariate_hypergeometric(colors, 6) array([5, 0, 1]) >>> gen.multivariate_hypergeometric(colors, 6, size=3) array([[5, 0, 1], [2, 2, 2], [3, 3, 0]]) >>> gen.multivariate_hypergeometric(colors, 6, size=(2, 2)) array([[[3, 2, 1], [3, 2, 1]], [[4, 1, 1], [3, 2, 1]]]) multivariate_hypergeometric multinomial(n, pvals, size=None) Draw samples from a multinomial distribution. The multinomial distribution is a multivariate generalization of the binomial distribution. Take an experiment with one of ``p`` possible outcomes. An example of such an experiment is throwing a dice, where the outcome can be 1 through 6. Each sample drawn from the distribution represents `n` such experiments. Its values, ``X_i = [X_0, X_1, ..., X_p]``, represent the number of times the outcome was ``i``. Parameters ---------- n : int or array-like of ints Number of experiments. pvals : array-like of floats Probabilities of each of the ``p`` different outcomes with shape ``(k0, k1, ..., kn, p)``. Each element ``pvals[i,j,...,:]`` must sum to 1 (however, the last element is always assumed to account for the remaining probability, as long as ``sum(pvals[..., :-1], axis=-1) <= 1.0``. Must have at least 1 dimension where pvals.shape[-1] > 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn each with ``p`` elements. Default is None where the output size is determined by the broadcast shape of ``n`` and all by the final dimension of ``pvals``, which is denoted as ``b=(b0, b1, ..., bq)``. If size is not None, then it must be compatible with the broadcast shape ``b``. Specifically, size must have ``q`` or more elements and size[-(q-j):] must equal ``bj``. Returns ------- out : ndarray The drawn samples, of shape size, if provided. When size is provided, the output shape is size + (p,) If not specified, the shape is determined by the broadcast shape of ``n`` and ``pvals``, ``(b0, b1, ..., bq)`` augmented with the dimension of the multinomial, ``p``, so that that output shape is ``(b0, b1, ..., bq, p)``. Each entry ``out[i,j,...,:]`` is a ``p``-dimensional value drawn from the distribution. Examples -------- Throw a dice 20 times: >>> rng = np.random.default_rng() >>> rng.multinomial(20, [1/6.]*6, size=1) array([[4, 1, 7, 5, 2, 1]]) # random It landed 4 times on 1, once on 2, etc. Now, throw the dice 20 times, and 20 times again: >>> rng.multinomial(20, [1/6.]*6, size=2) array([[3, 4, 3, 3, 4, 3], [2, 4, 3, 4, 0, 7]]) # random For the first run, we threw 3 times 1, 4 times 2, etc. For the second, we threw 2 times 1, 4 times 2, etc. Now, do one experiment throwing the dice 10 time, and 10 times again, and another throwing the dice 20 times, and 20 times again: >>> rng.multinomial([[10], [20]], [1/6.]*6, size=(2, 2)) array([[[2, 4, 0, 1, 2, 1], [1, 3, 0, 3, 1, 2]], [[1, 4, 4, 4, 4, 3], [3, 3, 2, 5, 5, 2]]]) # random The first array shows the outcomes of throwing the dice 10 times, and the second shows the outcomes from throwing the dice 20 times. A loaded die is more likely to land on number 6: >>> rng.multinomial(100, [1/7.]*5 + [2/7.]) array([11, 16, 14, 17, 16, 26]) # random Simulate 10 throws of a 4-sided die and 20 throws of a 6-sided die >>> rng.multinomial([10, 20],[[1/4]*4 + [0]*2, [1/6]*6]) array([[2, 1, 4, 3, 0, 0], [3, 3, 3, 6, 1, 4]], dtype=int64) # random Generate categorical random variates from two categories where the first has 3 outcomes and the second has 2. >>> rng.multinomial(1, [[.1, .5, .4 ], [.3, .7, .0]]) array([[0, 0, 1], [0, 1, 0]], dtype=int64) # random ``argmax(axis=-1)`` is then used to return the categories. >>> pvals = [[.1, .5, .4 ], [.3, .7, .0]] >>> rvs = rng.multinomial(1, pvals, size=(4,2)) >>> rvs.argmax(axis=-1) array([[0, 1], [2, 0], [2, 1], [2, 0]], dtype=int64) # random The same output dimension can be produced using broadcasting. >>> rvs = rng.multinomial([[1]] * 4, pvals) >>> rvs.argmax(axis=-1) array([[0, 1], [2, 0], [2, 1], [2, 0]], dtype=int64) # random The probability inputs should be normalized. As an implementation detail, the value of the last entry is ignored and assumed to take up any leftover probability mass, but this should not be relied on. A biased coin which has twice as much weight on one side as on the other should be sampled like so: >>> rng.multinomial(100, [1.0 / 3, 2.0 / 3]) # RIGHT array([38, 62]) # random not like: >>> rng.multinomial(100, [1.0, 2.0]) # WRONG Traceback (most recent call last): ValueError: pvals < 0, pvals > 1 or pvals contains NaNs multinomialmultinmu_mt19937msg__module__mode > rightmodemnixmnarrmethod must be one of {'eigh', 'svd', 'cholesky'}method must be "count" or "marginals".methodmemviewmemory allocation failed in permutedmean must be 1 dimensionalmean and cov must not be complexmean and cov must have same lengthmeanmay_share_memorymax_lam_arrmax_indexmax_maskedmaskmarginals__main___lowlow logseries(p, size=None) Draw samples from a logarithmic series distribution. Samples are drawn from a log series distribution with specified shape parameter, 0 <= ``p`` < 1. Parameters ---------- p : float or array_like of floats Shape parameter for the distribution. Must be in the range [0, 1). size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``p`` is a scalar. Otherwise, ``np.array(p).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized logarithmic series distribution. See Also -------- scipy.stats.logser : probability density function, distribution or cumulative density function, etc. Notes ----- The probability mass function for the Log Series distribution is .. math:: P(k) = \frac{-p^k}{k \ln(1-p)}, where p = probability. The log series distribution is frequently used to represent species richness and occurrence, first proposed by Fisher, Corbet, and Williams in 1943 [2]. It may also be used to model the numbers of occupants seen in cars [3]. References ---------- .. [1] Buzas, Martin A.; Culver, Stephen J., Understanding regional species diversity through the log series distribution of occurrences: BIODIVERSITY RESEARCH Diversity & Distributions, Volume 5, Number 5, September 1999 , pp. 187-195(9). .. [2] Fisher, R.A,, A.S. Corbet, and C.B. Williams. 1943. The relation between the number of species and the number of individuals in a random sample of an animal population. Journal of Animal Ecology, 12:42-58. .. [3] D. J. Hand, F. Daly, D. Lunn, E. Ostrowski, A Handbook of Small Data Sets, CRC Press, 1994. .. [4] Wikipedia, "Logarithmic distribution", https://en.wikipedia.org/wiki/Logarithmic_distribution Examples -------- Draw samples from the distribution: >>> a = .6 >>> rng = np.random.default_rng() >>> s = rng.logseries(a, 10000) >>> import matplotlib.pyplot as plt >>> bins = np.arange(-.5, max(s) + .5 ) >>> count, bins, _ = plt.hist(s, bins=bins, label='Sample count') # plot against distribution >>> def logseries(k, p): ... return -p**k/(k*np.log(1-p)) >>> centres = np.arange(1, max(s) + 1) >>> plt.plot(centres, logseries(centres, a) * s.size, 'r', label='logseries PMF') >>> plt.legend() >>> plt.show() logseries lognormal(mean=0.0, sigma=1.0, size=None) Draw samples from a log-normal distribution. Draw samples from a log-normal distribution with specified mean, standard deviation, and array shape. Note that the mean and standard deviation are not the values for the distribution itself, but of the underlying normal distribution it is derived from. Parameters ---------- mean : float or array_like of floats, optional Mean value of the underlying normal distribution. Default is 0. sigma : float or array_like of floats, optional Standard deviation of the underlying normal distribution. Must be non-negative. Default is 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``mean`` and ``sigma`` are both scalars. Otherwise, ``np.broadcast(mean, sigma).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized log-normal distribution. See Also -------- scipy.stats.lognorm : probability density function, distribution, cumulative density function, etc. Notes ----- A variable `x` has a log-normal distribution if `log(x)` is normally distributed. The probability density function for the log-normal distribution is: .. math:: p(x) = \frac{1}{\sigma x \sqrt{2\pi}} e^{(-\frac{(ln(x)-\mu)^2}{2\sigma^2})} where :math:`\mu` is the mean and :math:`\sigma` is the standard deviation of the normally distributed logarithm of the variable. A log-normal distribution results if a random variable is the *product* of a large number of independent, identically-distributed variables in the same way that a normal distribution results if the variable is the *sum* of a large number of independent, identically-distributed variables. References ---------- .. [1] Limpert, E., Stahel, W. A., and Abbt, M., "Log-normal Distributions across the Sciences: Keys and Clues," BioScience, Vol. 51, No. 5, May, 2001. https://stat.ethz.ch/~stahel/lognormal/bioscience.pdf .. [2] Reiss, R.D. and Thomas, M., "Statistical Analysis of Extreme Values," Basel: Birkhauser Verlag, 2001, pp. 31-32. Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> mu, sigma = 3., 1. # mean and standard deviation >>> s = rng.lognormal(mu, sigma, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, _ = plt.hist(s, 100, density=True, align='mid') >>> x = np.linspace(min(bins), max(bins), 10000) >>> pdf = (np.exp(-(np.log(x) - mu)**2 / (2 * sigma**2)) ... / (x * sigma * np.sqrt(2 * np.pi))) >>> plt.plot(x, pdf, linewidth=2, color='r') >>> plt.axis('tight') >>> plt.show() Demonstrate that taking the products of random samples from a uniform distribution can be fit well by a log-normal probability density function. >>> # Generate a thousand samples: each is the product of 100 random >>> # values, drawn from a normal distribution. >>> rng = rng >>> b = [] >>> for i in range(1000): ... a = 10. + rng.standard_normal(100) ... b.append(np.prod(a)) >>> b = np.array(b) / np.min(b) # scale values to be positive >>> count, bins, _ = plt.hist(b, 100, density=True, align='mid') >>> sigma = np.std(np.log(b)) >>> mu = np.mean(np.log(b)) >>> x = np.linspace(min(bins), max(bins), 10000) >>> pdf = (np.exp(-(np.log(x) - mu)**2 / (2 * sigma**2)) ... / (x * sigma * np.sqrt(2 * np.pi))) >>> plt.plot(x, pdf, color='r', linewidth=2) >>> plt.show() lognormal logistic(loc=0.0, scale=1.0, size=None) Draw samples from a logistic distribution. Samples are drawn from a logistic distribution with specified parameters, loc (location or mean, also median), and scale (>0). Parameters ---------- loc : float or array_like of floats, optional Parameter of the distribution. Default is 0. scale : float or array_like of floats, optional Parameter of the distribution. Must be non-negative. Default is 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``loc`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(loc, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized logistic distribution. See Also -------- scipy.stats.logistic : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Logistic distribution is .. math:: P(x) = \frac{e^{-(x-\mu)/s}}{s(1+e^{-(x-\mu)/s})^2}, where :math:`\mu` = location and :math:`s` = scale. The Logistic distribution is used in Extreme Value problems where it can act as a mixture of Gumbel distributions, in Epidemiology, and by the World Chess Federation (FIDE) where it is used in the Elo ranking system, assuming the performance of each player is a logistically distributed random variable. References ---------- .. [1] Reiss, R.-D. and Thomas M. (2001), "Statistical Analysis of Extreme Values, from Insurance, Finance, Hydrology and Other Fields," Birkhauser Verlag, Basel, pp 132-133. .. [2] Weisstein, Eric W. "Logistic Distribution." From MathWorld--A Wolfram Web Resource. https://mathworld.wolfram.com/LogisticDistribution.html .. [3] Wikipedia, "Logistic-distribution", https://en.wikipedia.org/wiki/Logistic_distribution Examples -------- Draw samples from the distribution: >>> loc, scale = 10, 1 >>> rng = np.random.default_rng() >>> s = rng.logistic(loc, scale, 10000) >>> import matplotlib.pyplot as plt >>> count, bins, _ = plt.hist(s, bins=50, label='Sampled data') # plot sampled data against the exact distribution >>> def logistic(x, loc, scale): ... return np.exp((loc-x)/scale)/(scale*(1+np.exp((loc-x)/scale))**2) >>> logistic_values = logistic(bins, loc, scale) >>> bin_spacing = np.mean(np.diff(bins)) >>> plt.plot(bins, logistic_values * bin_spacing * s.size, label='Logistic PDF') >>> plt.legend() >>> plt.show() logisticlogical_orlockloclnsamplelngoodlnbadlesslengthleft == rightleft > modeleft laplace(loc=0.0, scale=1.0, size=None) Draw samples from the Laplace or double exponential distribution with specified location (or mean) and scale (decay). The Laplace distribution is similar to the Gaussian/normal distribution, but is sharper at the peak and has fatter tails. It represents the difference between two independent, identically distributed exponential random variables. Parameters ---------- loc : float or array_like of floats, optional The position, :math:`\mu`, of the distribution peak. Default is 0. scale : float or array_like of floats, optional :math:`\lambda`, the exponential decay. Default is 1. Must be non- negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``loc`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(loc, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Laplace distribution. Notes ----- It has the probability density function .. math:: f(x; \mu, \lambda) = \frac{1}{2\lambda} \exp\left(-\frac{|x - \mu|}{\lambda}\right). The first law of Laplace, from 1774, states that the frequency of an error can be expressed as an exponential function of the absolute magnitude of the error, which leads to the Laplace distribution. For many problems in economics and health sciences, this distribution seems to model the data better than the standard Gaussian distribution. References ---------- .. [1] Abramowitz, M. and Stegun, I. A. (Eds.). "Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables, 9th printing," New York: Dover, 1972. .. [2] Kotz, Samuel, et. al. "The Laplace Distribution and Generalizations, " Birkhauser, 2001. .. [3] Weisstein, Eric W. "Laplace Distribution." From MathWorld--A Wolfram Web Resource. https://mathworld.wolfram.com/LaplaceDistribution.html .. [4] Wikipedia, "Laplace distribution", https://en.wikipedia.org/wiki/Laplace_distribution Examples -------- Draw samples from the distribution >>> loc, scale = 0., 1. >>> rng = np.random.default_rng() >>> s = rng.laplace(loc, scale, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, _ = plt.hist(s, 30, density=True) >>> x = np.arange(-8., 8., .01) >>> pdf = np.exp(-abs(x-loc)/scale)/(2.*scale) >>> plt.plot(x, pdf) Plot Gaussian for comparison: >>> g = (1/(scale * np.sqrt(2 * np.pi)) * ... np.exp(-(x - loc)**2 / (2 * scale**2))) >>> plt.plot(x,g) laplacelamlkappakjitemsize <= 0 for cython.arrayitemsizeitemitissubdtypeisscalarisnativeisnanisfiniteisenabledis_scalar is not compatible with broadcast dimensions of inputs _is_coroutineinvalid_colorsinvaccintp integers(low, high=None, size=None, dtype=np.int64, endpoint=False) Return random integers from `low` (inclusive) to `high` (exclusive), or if endpoint=True, `low` (inclusive) to `high` (inclusive). Replaces `RandomState.randint` (with endpoint=False) and `RandomState.random_integers` (with endpoint=True) Return random integers from the "discrete uniform" distribution of the specified dtype. If `high` is None (the default), then results are from 0 to `low`. Parameters ---------- low : int or array-like of ints Lowest (signed) integers to be drawn from the distribution (unless ``high=None``, in which case this parameter is 0 and this value is used for `high`). high : int or array-like of ints, optional If provided, one above the largest (signed) integer to be drawn from the distribution (see above for behavior if ``high=None``). If array-like, must contain integer values size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. dtype : dtype, optional Desired dtype of the result. Byteorder must be native. The default value is np.int64. endpoint : bool, optional If true, sample from the interval [low, high] instead of the default [low, high) Defaults to False Returns ------- out : int or ndarray of ints `size`-shaped array of random integers from the appropriate distribution, or a single such random int if `size` not provided. Notes ----- When using broadcasting with uint64 dtypes, the maximum value (2**64) cannot be represented as a standard integer type. The high array (or low if high is None) must have object dtype, e.g., array([2**64]). Examples -------- >>> rng = np.random.default_rng() >>> rng.integers(2, size=10) array([1, 0, 0, 0, 1, 1, 0, 0, 1, 0]) # random >>> rng.integers(1, size=10) array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0]) Generate a 2 x 4 array of ints between 0 and 4, inclusive: >>> rng.integers(5, size=(2, 4)) array([[4, 0, 2, 1], [3, 2, 2, 0]]) # random Generate a 1 x 3 array with 3 different upper bounds >>> rng.integers(1, [3, 5, 10]) array([2, 2, 9]) # random Generate a 1 by 3 array with 3 different lower bounds >>> rng.integers([1, 5, 7], 10) array([9, 8, 7]) # random Generate a 2 by 4 array using broadcasting with dtype of uint8 >>> rng.integers([1, 3, 5, 7], [[10], [20]], dtype=np.uint8) array([[ 8, 6, 9, 7], [ 1, 16, 9, 12]], dtype=uint8) # random References ---------- .. [1] Daniel Lemire., "Fast Random Integer Generation in an Interval", ACM Transactions on Modeling and Computer Simulation 29 (1), 2019, https://arxiv.org/abs/1805.10941. integersint8int64int32int16_initializingindex_in__import__ignoreidx_dataidxidi hypergeometric(ngood, nbad, nsample, size=None) Draw samples from a Hypergeometric distribution. Samples are drawn from a hypergeometric distribution with specified parameters, `ngood` (ways to make a good selection), `nbad` (ways to make a bad selection), and `nsample` (number of items sampled, which is less than or equal to the sum ``ngood + nbad``). Parameters ---------- ngood : int or array_like of ints Number of ways to make a good selection. Must be nonnegative and less than 10**9. nbad : int or array_like of ints Number of ways to make a bad selection. Must be nonnegative and less than 10**9. nsample : int or array_like of ints Number of items sampled. Must be nonnegative and less than ``ngood + nbad``. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if `ngood`, `nbad`, and `nsample` are all scalars. Otherwise, ``np.broadcast(ngood, nbad, nsample).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized hypergeometric distribution. Each sample is the number of good items within a randomly selected subset of size `nsample` taken from a set of `ngood` good items and `nbad` bad items. See Also -------- multivariate_hypergeometric : Draw samples from the multivariate hypergeometric distribution. scipy.stats.hypergeom : probability density function, distribution or cumulative density function, etc. Notes ----- The probability mass function (PMF) for the Hypergeometric distribution is .. math:: P(x) = \frac{\binom{g}{x}\binom{b}{n-x}}{\binom{g+b}{n}}, where :math:`0 \le x \le n` and :math:`n-b \le x \le g` for P(x) the probability of ``x`` good results in the drawn sample, g = `ngood`, b = `nbad`, and n = `nsample`. Consider an urn with black and white marbles in it, `ngood` of them are black and `nbad` are white. If you draw `nsample` balls without replacement, then the hypergeometric distribution describes the distribution of black balls in the drawn sample. Note that this distribution is very similar to the binomial distribution, except that in this case, samples are drawn without replacement, whereas in the Binomial case samples are drawn with replacement (or the sample space is infinite). As the sample space becomes large, this distribution approaches the binomial. The arguments `ngood` and `nbad` each must be less than `10**9`. For extremely large arguments, the algorithm that is used to compute the samples [4]_ breaks down because of loss of precision in floating point calculations. For such large values, if `nsample` is not also large, the distribution can be approximated with the binomial distribution, `binomial(n=nsample, p=ngood/(ngood + nbad))`. References ---------- .. [1] Lentner, Marvin, "Elementary Applied Statistics", Bogden and Quigley, 1972. .. [2] Weisstein, Eric W. "Hypergeometric Distribution." From MathWorld--A Wolfram Web Resource. https://mathworld.wolfram.com/HypergeometricDistribution.html .. [3] Wikipedia, "Hypergeometric distribution", https://en.wikipedia.org/wiki/Hypergeometric_distribution .. [4] Stadlober, Ernst, "The ratio of uniforms approach for generating discrete random variates", Journal of Computational and Applied Mathematics, 31, pp. 181-189 (1990). Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> ngood, nbad, nsamp = 100, 2, 10 # number of good, number of bad, and number of samples >>> s = rng.hypergeometric(ngood, nbad, nsamp, 1000) >>> from matplotlib.pyplot import hist >>> hist(s) # note that it is very unlikely to grab both bad items Suppose you have an urn with 15 white and 15 black marbles. If you pull 15 marbles at random, how likely is it that 12 or more of them are one color? >>> s = rng.hypergeometric(15, 15, 15, 100000) >>> sum(s>=12)/100000. + sum(s<=3)/100000. # answer = 0.003 ... pretty unlikely! hypergeometrichigh - low range exceeds valid boundshigh - low_highhighhasobjecthash_set gumbel(loc=0.0, scale=1.0, size=None) Draw samples from a Gumbel distribution. Draw samples from a Gumbel distribution with specified location and scale. For more information on the Gumbel distribution, see Notes and References below. Parameters ---------- loc : float or array_like of floats, optional The location of the mode of the distribution. Default is 0. scale : float or array_like of floats, optional The scale parameter of the distribution. Default is 1. Must be non- negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``loc`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(loc, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Gumbel distribution. See Also -------- scipy.stats.gumbel_l scipy.stats.gumbel_r scipy.stats.genextreme weibull Notes ----- The Gumbel (or Smallest Extreme Value (SEV) or the Smallest Extreme Value Type I) distribution is one of a class of Generalized Extreme Value (GEV) distributions used in modeling extreme value problems. The Gumbel is a special case of the Extreme Value Type I distribution for maximums from distributions with "exponential-like" tails. The probability density for the Gumbel distribution is .. math:: p(x) = \frac{e^{-(x - \mu)/ \beta}}{\beta} e^{ -e^{-(x - \mu)/ \beta}}, where :math:`\mu` is the mode, a location parameter, and :math:`\beta` is the scale parameter. The Gumbel (named for German mathematician Emil Julius Gumbel) was used very early in the hydrology literature, for modeling the occurrence of flood events. It is also used for modeling maximum wind speed and rainfall rates. It is a "fat-tailed" distribution - the probability of an event in the tail of the distribution is larger than if one used a Gaussian, hence the surprisingly frequent occurrence of 100-year floods. Floods were initially modeled as a Gaussian process, which underestimated the frequency of extreme events. It is one of a class of extreme value distributions, the Generalized Extreme Value (GEV) distributions, which also includes the Weibull and Frechet. The function has a mean of :math:`\mu + 0.57721\beta` and a variance of :math:`\frac{\pi^2}{6}\beta^2`. References ---------- .. [1] Gumbel, E. J., "Statistics of Extremes," New York: Columbia University Press, 1958. .. [2] Reiss, R.-D. and Thomas, M., "Statistical Analysis of Extreme Values from Insurance, Finance, Hydrology and Other Fields," Basel: Birkhauser Verlag, 2001. Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> mu, beta = 0, 0.1 # location and scale >>> s = rng.gumbel(mu, beta, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, _ = plt.hist(s, 30, density=True) >>> plt.plot(bins, (1/beta)*np.exp(-(bins - mu)/beta) ... * np.exp( -np.exp( -(bins - mu) /beta) ), ... linewidth=2, color='r') >>> plt.show() Show how an extreme value distribution can arise from a Gaussian process and compare to a Gaussian: >>> means = [] >>> maxima = [] >>> for i in range(0,1000) : ... a = rng.normal(mu, beta, 1000) ... means.append(a.mean()) ... maxima.append(a.max()) >>> count, bins, _ = plt.hist(maxima, 30, density=True) >>> beta = np.std(maxima) * np.sqrt(6) / np.pi >>> mu = np.mean(maxima) - 0.57721*beta >>> plt.plot(bins, (1/beta)*np.exp(-(bins - mu)/beta) ... * np.exp(-np.exp(-(bins - mu)/beta)), ... linewidth=2, color='r') >>> plt.plot(bins, 1/(beta * np.sqrt(2 * np.pi)) ... * np.exp(-(bins - mu)**2 / (2 * beta**2)), ... linewidth=2, color='g') >>> plt.show() gumbelgreatergot differing extents in dimension (got __getstate__ geometric(p, size=None) Draw samples from the geometric distribution. Bernoulli trials are experiments with one of two outcomes: success or failure (an example of such an experiment is flipping a coin). The geometric distribution models the number of trials that must be run in order to achieve success. It is therefore supported on the positive integers, ``k = 1, 2, ...``. The probability mass function of the geometric distribution is .. math:: f(k) = (1 - p)^{k - 1} p where `p` is the probability of success of an individual trial. Parameters ---------- p : float or array_like of floats The probability of success of an individual trial. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``p`` is a scalar. Otherwise, ``np.array(p).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized geometric distribution. References ---------- .. [1] Wikipedia, "Geometric distribution", https://en.wikipedia.org/wiki/Geometric_distribution Examples -------- Draw 10,000 values from the geometric distribution, with the probability of an individual success equal to ``p = 0.35``: >>> p, size = 0.35, 10000 >>> rng = np.random.default_rng() >>> sample = rng.geometric(p=p, size=size) What proportion of trials succeeded after a single run? >>> (sample == 1).sum()/size 0.34889999999999999 # may vary The geometric distribution with ``p=0.35`` looks as follows: >>> import matplotlib.pyplot as plt >>> count, bins, _ = plt.hist(sample, bins=30, density=True) >>> plt.plot(bins, (1-p)**(bins-1)*p) >>> plt.xlim([0, 25]) >>> plt.show() geometric__generator_ctorgengc gamma(shape, scale=1.0, size=None) Draw samples from a Gamma distribution. Samples are drawn from a Gamma distribution with specified parameters, `shape` (sometimes designated "k") and `scale` (sometimes designated "theta"), where both parameters are > 0. Parameters ---------- shape : float or array_like of floats The shape of the gamma distribution. Must be non-negative. scale : float or array_like of floats, optional The scale of the gamma distribution. Must be non-negative. Default is equal to 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``shape`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(shape, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized gamma distribution. See Also -------- scipy.stats.gamma : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Gamma distribution is .. math:: p(x) = x^{k-1}\frac{e^{-x/\theta}}{\theta^k\Gamma(k)}, where :math:`k` is the shape and :math:`\theta` the scale, and :math:`\Gamma` is the Gamma function. The Gamma distribution is often used to model the times to failure of electronic components, and arises naturally in processes for which the waiting times between Poisson distributed events are relevant. References ---------- .. [1] Weisstein, Eric W. "Gamma Distribution." From MathWorld--A Wolfram Web Resource. https://mathworld.wolfram.com/GammaDistribution.html .. [2] Wikipedia, "Gamma distribution", https://en.wikipedia.org/wiki/Gamma_distribution Examples -------- Draw samples from the distribution: >>> shape, scale = 2., 2. # mean=4, std=2*sqrt(2) >>> rng = np.random.default_rng() >>> s = rng.gamma(shape, scale, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> import scipy.special as sps # doctest: +SKIP >>> count, bins, _ = plt.hist(s, 50, density=True) >>> y = bins**(shape-1)*(np.exp(-bins/scale) / # doctest: +SKIP ... (sps.gamma(shape)*scale**shape)) >>> plt.plot(bins, y, linewidth=2, color='r') # doctest: +SKIP >>> plt.show() gammagfunc__func__fullfrightfoundfortranformatfmodefloat64float32fleftflat_foundflagsfinfofinal_shape_factor f(dfnum, dfden, size=None) Draw samples from an F distribution. Samples are drawn from an F distribution with specified parameters, `dfnum` (degrees of freedom in numerator) and `dfden` (degrees of freedom in denominator), where both parameters must be greater than zero. The random variate of the F distribution (also known as the Fisher distribution) is a continuous probability distribution that arises in ANOVA tests, and is the ratio of two chi-square variates. Parameters ---------- dfnum : float or array_like of floats Degrees of freedom in numerator, must be > 0. dfden : float or array_like of float Degrees of freedom in denominator, must be > 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``dfnum`` and ``dfden`` are both scalars. Otherwise, ``np.broadcast(dfnum, dfden).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Fisher distribution. See Also -------- scipy.stats.f : probability density function, distribution or cumulative density function, etc. Notes ----- The F statistic is used to compare in-group variances to between-group variances. Calculating the distribution depends on the sampling, and so it is a function of the respective degrees of freedom in the problem. The variable `dfnum` is the number of samples minus one, the between-groups degrees of freedom, while `dfden` is the within-groups degrees of freedom, the sum of the number of samples in each group minus the number of groups. References ---------- .. [1] Glantz, Stanton A. "Primer of Biostatistics.", McGraw-Hill, Fifth Edition, 2002. .. [2] Wikipedia, "F-distribution", https://en.wikipedia.org/wiki/F-distribution Examples -------- An example from Glantz[1], pp 47-40: Two groups, children of diabetics (25 people) and children from people without diabetes (25 controls). Fasting blood glucose was measured, case group had a mean value of 86.1, controls had a mean value of 82.2. Standard deviations were 2.09 and 2.49 respectively. Are these data consistent with the null hypothesis that the parents diabetic status does not affect their children's blood glucose levels? Calculating the F statistic from the data gives a value of 36.01. Draw samples from the distribution: >>> dfnum = 1. # between group degrees of freedom >>> dfden = 48. # within groups degrees of freedom >>> rng = np.random.default_rng() >>> s = rng.f(dfnum, dfden, 1000) The lower bound for the top 1% of the samples is : >>> np.sort(s)[-10] 7.61988120985 # random So there is about a 1% chance that the F statistic will exceed 7.62, the measured value is 36, so the null hypothesis is rejected at the 1% level. The corresponding probability density function for ``n = 20`` and ``m = 20`` is: >>> import matplotlib.pyplot as plt >>> from scipy import stats >>> dfnum, dfden, size = 20, 20, 10000 >>> s = rng.f(dfnum=dfnum, dfden=dfden, size=size) >>> bins, density, _ = plt.hist(s, 30, density=True) >>> x = np.linspace(0, 5, 1000) >>> plt.plot(x, stats.f.pdf(x, dfnum, dfden)) >>> plt.xlim([0, 5]) >>> plt.show() f exponential(scale=1.0, size=None) Draw samples from an exponential distribution. Its probability density function is .. math:: f(x; \frac{1}{\beta}) = \frac{1}{\beta} \exp(-\frac{x}{\beta}), for ``x > 0`` and 0 elsewhere. :math:`\beta` is the scale parameter, which is the inverse of the rate parameter :math:`\lambda = 1/\beta`. The rate parameter is an alternative, widely used parameterization of the exponential distribution [3]_. The exponential distribution is a continuous analogue of the geometric distribution. It describes many common situations, such as the size of raindrops measured over many rainstorms [1]_, or the time between page requests to Wikipedia [2]_. Parameters ---------- scale : float or array_like of floats The scale parameter, :math:`\beta = 1/\lambda`. Must be non-negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``scale`` is a scalar. Otherwise, ``np.array(scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized exponential distribution. Examples -------- Assume a company has 10000 customer support agents and the time between customer calls is exponentially distributed and that the average time between customer calls is 4 minutes. >>> scale, size = 4, 10000 >>> rng = np.random.default_rng() >>> time_between_calls = rng.exponential(scale=scale, size=size) What is the probability that a customer will call in the next 4 to 5 minutes? >>> x = ((time_between_calls < 5).sum())/size >>> y = ((time_between_calls < 4).sum())/size >>> x - y 0.08 # may vary The corresponding distribution can be visualized as follows: >>> import matplotlib.pyplot as plt >>> scale, size = 4, 10000 >>> rng = np.random.default_rng() >>> sample = rng.exponential(scale=scale, size=size) >>> count, bins, _ = plt.hist(sample, 30, density=True) >>> plt.plot(bins, scale**(-1)*np.exp(-scale**-1*bins), linewidth=2, color='r') >>> plt.show() References ---------- .. [1] Peyton Z. Peebles Jr., "Probability, Random Variables and Random Signal Principles", 4th ed, 2001, p. 57. .. [2] Wikipedia, "Poisson process", https://en.wikipedia.org/wiki/Poisson_process .. [3] Wikipedia, "Exponential distribution", https://en.wikipedia.org/wiki/Exponential_distribution exponential__exit__excerrorequalepsenumerate__enter__endpointencodeenableempty_likeemptyeighdtype_is_object_dtypedtype_dpdoubledot_dn_dmax_lamdisable dirichlet(alpha, size=None) Draw samples from the Dirichlet distribution. Draw `size` samples of dimension k from a Dirichlet distribution. A Dirichlet-distributed random variable can be seen as a multivariate generalization of a Beta distribution. The Dirichlet distribution is a conjugate prior of a multinomial distribution in Bayesian inference. Parameters ---------- alpha : sequence of floats, length k Parameter of the distribution (length ``k`` for sample of length ``k``). size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n)``, then ``m * n * k`` samples are drawn. Default is None, in which case a vector of length ``k`` is returned. Returns ------- samples : ndarray, The drawn samples, of shape ``(size, k)``. Raises ------ ValueError If any value in ``alpha`` is less than zero Notes ----- The Dirichlet distribution is a distribution over vectors :math:`x` that fulfil the conditions :math:`x_i>0` and :math:`\sum_{i=1}^k x_i = 1`. The probability density function :math:`p` of a Dirichlet-distributed random vector :math:`X` is proportional to .. math:: p(x) \propto \prod_{i=1}^{k}{x^{\alpha_i-1}_i}, where :math:`\alpha` is a vector containing the positive concentration parameters. The method uses the following property for computation: let :math:`Y` be a random vector which has components that follow a standard gamma distribution, then :math:`X = \frac{1}{\sum_{i=1}^k{Y_i}} Y` is Dirichlet-distributed References ---------- .. [1] David McKay, "Information Theory, Inference and Learning Algorithms," chapter 23, https://www.inference.org.uk/mackay/itila/ .. [2] Wikipedia, "Dirichlet distribution", https://en.wikipedia.org/wiki/Dirichlet_distribution Examples -------- Taking an example cited in Wikipedia, this distribution can be used if one wanted to cut strings (each of initial length 1.0) into K pieces with different lengths, where each piece had, on average, a designated average length, but allowing some variation in the relative sizes of the pieces. >>> rng = np.random.default_rng() >>> s = rng.dirichlet((10, 5, 3), 20).transpose() >>> import matplotlib.pyplot as plt >>> plt.barh(range(20), s[0]) >>> plt.barh(range(20), s[1], left=s[0], color='g') >>> plt.barh(range(20), s[2], left=s[0]+s[1], color='r') >>> plt.title("Lengths of Strings") dirichletdiric__dict__dfnumdfdendfdefault_rng (line 4989)default_rngdcutoffcumsumcsumcovariance is not symmetric positive-semidefinite.cov must be 2 dimensional and squarecovcount_nonzerocountcopytocopycolors_ptrcolors must be a one-dimensional sequence of nonnegative integers not exceeding %d.colorscollections.abccntcline_in_traceback__class_getitem____class__cholesky choice(a, size=None, replace=True, p=None, axis=0, shuffle=True) Generates a random sample from a given array Parameters ---------- a : {array_like, int} If an ndarray, a random sample is generated from its elements. If an int, the random sample is generated from np.arange(a). size : {int, tuple[int]}, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn from the 1-d `a`. If `a` has more than one dimension, the `size` shape will be inserted into the `axis` dimension, so the output ``ndim`` will be ``a.ndim - 1 + len(size)``. Default is None, in which case a single value is returned. replace : bool, optional Whether the sample is with or without replacement. Default is True, meaning that a value of ``a`` can be selected multiple times. p : 1-D array_like, optional The probabilities associated with each entry in a. If not given, the sample assumes a uniform distribution over all entries in ``a``. axis : int, optional The axis along which the selection is performed. The default, 0, selects by row. shuffle : bool, optional Whether the sample is shuffled when sampling without replacement. Default is True, False provides a speedup. Returns ------- samples : single item or ndarray The generated random samples Raises ------ ValueError If a is an int and less than zero, if p is not 1-dimensional, if a is array-like with a size 0, if p is not a vector of probabilities, if a and p have different lengths, or if replace=False and the sample size is greater than the population size. See Also -------- integers, shuffle, permutation Notes ----- Setting user-specified probabilities through ``p`` uses a more general but less efficient sampler than the default. The general sampler produces a different sample than the optimized sampler even if each element of ``p`` is 1 / len(a). ``p`` must sum to 1 when cast to ``float64``. To ensure this, you may wish to normalize using ``p = p / np.sum(p, dtype=float)``. When passing ``a`` as an integer type and ``size`` is not specified, the return type is a native Python ``int``. Examples -------- Generate a uniform random sample from np.arange(5) of size 3: >>> rng = np.random.default_rng() >>> rng.choice(5, 3) array([0, 3, 4]) # random >>> #This is equivalent to rng.integers(0,5,3) Generate a non-uniform random sample from np.arange(5) of size 3: >>> rng.choice(5, 3, p=[0.1, 0, 0.3, 0.6, 0]) array([3, 3, 0]) # random Generate a uniform random sample from np.arange(5) of size 3 without replacement: >>> rng.choice(5, 3, replace=False) array([3,1,0]) # random >>> #This is equivalent to rng.permutation(np.arange(5))[:3] Generate a uniform random sample from a 2-D array along the first axis (the default), without replacement: >>> rng.choice([[0, 1, 2], [3, 4, 5], [6, 7, 8]], 2, replace=False) array([[3, 4, 5], # random [0, 1, 2]]) Generate a non-uniform random sample from np.arange(5) of size 3 without replacement: >>> rng.choice(5, 3, replace=False, p=[0.1, 0, 0.3, 0.6, 0]) array([2, 3, 0]) # random Any of the above can be repeated with an arbitrary array-like instead of just integers. For instance: >>> aa_milne_arr = ['pooh', 'rabbit', 'piglet', 'Christopher'] >>> rng.choice(aa_milne_arr, 5, p=[0.5, 0.1, 0.1, 0.3]) array(['pooh', 'pooh', 'pooh', 'Christopher', 'piglet'], # random dtype=' 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``df`` is a scalar. Otherwise, ``np.array(df).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized chi-square distribution. Raises ------ ValueError When `df` <= 0 or when an inappropriate `size` (e.g. ``size=-1``) is given. Notes ----- The variable obtained by summing the squares of `df` independent, standard normally distributed random variables: .. math:: Q = \sum_{i=1}^{\mathtt{df}} X^2_i is chi-square distributed, denoted .. math:: Q \sim \chi^2_k. The probability density function of the chi-squared distribution is .. math:: p(x) = \frac{(1/2)^{k/2}}{\Gamma(k/2)} x^{k/2 - 1} e^{-x/2}, where :math:`\Gamma` is the gamma function, .. math:: \Gamma(x) = \int_0^{-\infty} t^{x - 1} e^{-t} dt. References ---------- .. [1] NIST "Engineering Statistics Handbook" https://www.itl.nist.gov/div898/handbook/eda/section3/eda3666.htm Examples -------- >>> rng = np.random.default_rng() >>> rng.chisquare(2,4) array([ 1.89920014, 9.00867716, 3.13710533, 5.62318272]) # random The distribution of a chi-square random variable with 20 degrees of freedom looks as follows: >>> import matplotlib.pyplot as plt >>> import scipy.stats as stats >>> s = rng.chisquare(20, 10000) >>> count, bins, _ = plt.hist(s, 30, density=True) >>> x = np.linspace(0, 60, 1000) >>> plt.plot(x, stats.chi2.pdf(x, df=20)) >>> plt.xlim([0, 60]) >>> plt.show() chisquarecheck_valid must equal 'warn', 'raise', or 'ignore'check_validcdfcastingcapsulec bytes(length) Return random bytes. Parameters ---------- length : int Number of random bytes. Returns ------- out : bytes String of length `length`. Notes ----- This function generates random bytes from a discrete uniform distribution. The generated bytes are independent from the CPU's native endianness. Examples -------- >>> rng = np.random.default_rng() >>> rng.bytes(10) b'\xfeC\x9b\x86\x17\xf2\xa1\xafcp' # random bytesbuf_ptrbufboth ngood and nbad must be less than %dbool_bit_generatorbit_generatorbit_gen binomial(n, p, size=None) Draw samples from a binomial distribution. Samples are drawn from a binomial distribution with specified parameters, n trials and p probability of success where n an integer >= 0 and p is in the interval [0,1]. (n may be input as a float, but it is truncated to an integer in use) Parameters ---------- n : int or array_like of ints Parameter of the distribution, >= 0. Floats are also accepted, but they will be truncated to integers. p : float or array_like of floats Parameter of the distribution, >= 0 and <=1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``n`` and ``p`` are both scalars. Otherwise, ``np.broadcast(n, p).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized binomial distribution, where each sample is equal to the number of successes over the n trials. See Also -------- scipy.stats.binom : probability density function, distribution or cumulative density function, etc. Notes ----- The probability mass function (PMF) for the binomial distribution is .. math:: P(N) = \binom{n}{N}p^N(1-p)^{n-N}, where :math:`n` is the number of trials, :math:`p` is the probability of success, and :math:`N` is the number of successes. When estimating the standard error of a proportion in a population by using a random sample, the normal distribution works well unless the product p*n <=5, where p = population proportion estimate, and n = number of samples, in which case the binomial distribution is used instead. For example, a sample of 15 people shows 4 who are left handed, and 11 who are right handed. Then p = 4/15 = 27%. 0.27*15 = 4, so the binomial distribution should be used in this case. References ---------- .. [1] Dalgaard, Peter, "Introductory Statistics with R", Springer-Verlag, 2002. .. [2] Glantz, Stanton A. "Primer of Biostatistics.", McGraw-Hill, Fifth Edition, 2002. .. [3] Lentner, Marvin, "Elementary Applied Statistics", Bogden and Quigley, 1972. .. [4] Weisstein, Eric W. "Binomial Distribution." From MathWorld--A Wolfram Web Resource. https://mathworld.wolfram.com/BinomialDistribution.html .. [5] Wikipedia, "Binomial distribution", https://en.wikipedia.org/wiki/Binomial_distribution Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> n, p, size = 10, .5, 10000 >>> s = rng.binomial(n, p, 10000) Assume a company drills 9 wild-cat oil exploration wells, each with an estimated probability of success of ``p=0.1``. All nine wells fail. What is the probability of that happening? Over ``size = 20,000`` trials the probability of this happening is on average: >>> n, p, size = 9, 0.1, 20000 >>> np.sum(rng.binomial(n=n, p=p, size=size) == 0)/size 0.39015 # may vary The following can be used to visualize a sample with ``n=100``, ``p=0.4`` and the corresponding probability density function: >>> import matplotlib.pyplot as plt >>> from scipy.stats import binom >>> n, p, size = 100, 0.4, 10000 >>> sample = rng.binomial(n, p, size=size) >>> count, bins, _ = plt.hist(sample, 30, density=True) >>> x = np.arange(n) >>> y = binom.pmf(x, n, p) >>> plt.plot(x, y, linewidth=2, color='r') binomial beta(a, b, size=None) Draw samples from a Beta distribution. The Beta distribution is a special case of the Dirichlet distribution, and is related to the Gamma distribution. It has the probability distribution function .. math:: f(x; a,b) = \frac{1}{B(\alpha, \beta)} x^{\alpha - 1} (1 - x)^{\beta - 1}, where the normalization, B, is the beta function, .. math:: B(\alpha, \beta) = \int_0^1 t^{\alpha - 1} (1 - t)^{\beta - 1} dt. It is often seen in Bayesian inference and order statistics. Parameters ---------- a : float or array_like of floats Alpha, positive (>0). b : float or array_like of floats Beta, positive (>0). size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``a`` and ``b`` are both scalars. Otherwise, ``np.broadcast(a, b).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized beta distribution. Examples -------- The beta distribution has mean a/(a+b). If ``a == b`` and both are > 1, the distribution is symmetric with mean 0.5. >>> rng = np.random.default_rng() >>> a, b, size = 2.0, 2.0, 10000 >>> sample = rng.beta(a=a, b=b, size=size) >>> np.mean(sample) 0.5047328775385895 # may vary Otherwise the distribution is skewed left or right according to whether ``a`` or ``b`` is greater. The distribution is mirror symmetric. See for example: >>> a, b, size = 2, 7, 10000 >>> sample_left = rng.beta(a=a, b=b, size=size) >>> sample_right = rng.beta(a=b, b=a, size=size) >>> m_left, m_right = np.mean(sample_left), np.mean(sample_right) >>> print(m_left, m_right) 0.2238596793678923 0.7774613834041182 # may vary >>> print(m_left - a/(a+b)) 0.001637457145670096 # may vary >>> print(m_right - b/(a+b)) -0.0003163943736596009 # may vary Display the histogram of the two samples: >>> import matplotlib.pyplot as plt >>> plt.hist([sample_left, sample_right], ... 50, density=True, histtype='bar') >>> plt.show() References ---------- .. [1] Wikipedia, "Beta distribution", https://en.wikipedia.org/wiki/Beta_distribution betabasebaxstrideaxlenaxisaxatol at 0xasyncio.coroutines.astype(np.float64)) > 1.0. The pvals array is cast to 64-bit floating point prior to checking the sum. Precision changes when casting may cause problems even if the sum of the original pvals is valid.astypeascontiguousarrayasarrayarray is read-onlyarrayarrarangeany and alpha_dataalpha_csum_dataalpha_csum_arralpha_arralpha < 0alphaalowallocate_bufferallcloseallahighadd_noteaddaccabca_originala must be a sequence or an integer, not a must be a positive integer unless no samples are takena cannot be empty unless no samples are takena and p must have same sizea()'>._?XWhen method is "marginals", sum(colors) must be less than 1000000000.When method is 'count', sum(colors) must not exceed %dView.MemoryViewValueErrorUserWarningUnsupported dtype %r for standard_gammaUnsupported dtype %r for standard_normalUnsupported dtype %r for standard_exponentialUnsupported dtype %r for randomUnsupported dtype %r for integersUnable to convert item to objectTypeErrorTStep may not be zero (axis %d)SequenceRuntimeWarningRange exceeds valid boundsRandomStateProviding a dtype with a non-native byteorder is not supported. If you require platform-independent byteorder, call byteswap when required.Probabilities do not sum to 1. See Notes section of docstring for more information.Probabilities contain NaNProbabilities are not non-negativePickleErrorPCG64OverflowErrorOutput size Out of bounds on buffer access (axis ONote that Cython is deliberately stricter than PEP-484 and rejects subclasses of builtin types. If you need to pass subclasses then set the 'annotation_typing' directive to False.NotImplementedErrorNone>> import numpy as np >>> rng = np.random.default_rng(12345) >>> print(rng) Generator(PCG64) >>> rfloat = rng.random() >>> rfloat 0.22733602246716966 >>> type(rfloat) Here we use `default_rng` to generate 3 random integers between 0 (inclusive) and 10 (exclusive): >>> import numpy as np >>> rng = np.random.default_rng(12345) >>> rints = rng.integers(low=0, high=10, size=3) >>> rints array([6, 2, 7]) >>> type(rints[0]) Here we specify a seed so that we have reproducible results: >>> import numpy as np >>> rng = np.random.default_rng(seed=42) >>> print(rng) Generator(PCG64) >>> arr1 = rng.random((3, 3)) >>> arr1 array([[0.77395605, 0.43887844, 0.85859792], [0.69736803, 0.09417735, 0.97562235], [0.7611397 , 0.78606431, 0.12811363]]) If we exit and restart our Python interpreter, we'll see that we generate the same random numbers again: >>> import numpy as np >>> rng = np.random.default_rng(seed=42) >>> arr2 = rng.random((3, 3)) >>> arr2 array([[0.77395605, 0.43887844, 0.85859792], [0.69736803, 0.09417735, 0.97562235], [0.7611397 , 0.78606431, 0.12811363]]) Cannot transpose memoryview with indirect dimensionsCannot take a larger sample than population when replace is FalseCannot index with type 'Cannot create writable memory view from read-only memoryviewCannot assign to read-only memoryviewCan only create a buffer that is contiguous in memory.Buffer view does not expose stridesAxis argument is only supported on ndarray objectsAssertionErrorAll dimensions preceding dimension %d must be indexed and not slicedASCIIA[...,:-1]) > 1.0[:-1]: default_rng(seed=None) Construct a new Generator with the default BitGenerator (PCG64). Parameters ---------- seed : {None, int, array_like[ints], SeedSequence, BitGenerator, Generator, RandomState}, optional A seed to initialize the `BitGenerator`. If None, then fresh, unpredictable entropy will be pulled from the OS. If an ``int`` or ``array_like[ints]`` is passed, then all values must be non-negative and will be passed to `SeedSequence` to derive the initial `BitGenerator` state. One may also pass in a `SeedSequence` instance. Additionally, when passed a `BitGenerator`, it will be wrapped by `Generator`. If passed a `Generator`, it will be returned unaltered. When passed a legacy `RandomState` instance it will be coerced to a `Generator`. Returns ------- Generator The initialized generator object. Notes ----- If ``seed`` is not a `BitGenerator` or a `Generator`, a new `BitGenerator` is instantiated. This function does not manage a default global instance. See :ref:`seeding_and_entropy` for more information about seeding. Examples -------- `default_rng` is the recommended constructor for the random number class `Generator`. Here are several ways we can construct a random number generator using `default_rng` and the `Generator` class. Here we use `default_rng` to generate a random float: >>> import numpy as np >>> rng = np.random.default_rng(12345) >>> print(rng) Generator(PCG64) >>> rfloat = rng.random() >>> rfloat 0.22733602246716966 >>> type(rfloat) Here we use `default_rng` to generate 3 random integers between 0 (inclusive) and 10 (exclusive): >>> import numpy as np >>> rng = np.random.default_rng(12345) >>> rints = rng.integers(low=0, high=10, size=3) >>> rints array([6, 2, 7]) >>> type(rints[0]) Here we specify a seed so that we have reproducible results: >>> import numpy as np >>> rng = np.random.default_rng(seed=42) >>> print(rng) Generator(PCG64) >>> arr1 = rng.random((3, 3)) >>> arr1 array([[0.77395605, 0.43887844, 0.85859792], [0.69736803, 0.09417735, 0.97562235], [0.7611397 , 0.78606431, 0.12811363]]) If we exit and restart our Python interpreter, we'll see that we generate the same random numbers again: >>> import numpy as np >>> rng = np.random.default_rng(seed=42) >>> arr2 = rng.random((3, 3)) >>> arr2 array([[0.77395605, 0.43887844, 0.85859792], [0.69736803, 0.09417735, 0.97562235], [0.7611397 , 0.78606431, 0.12811363]]) permutation(x, axis=0) Randomly permute a sequence, or return a permuted range. Parameters ---------- x : int or array_like If `x` is an integer, randomly permute ``np.arange(x)``. If `x` is an array, make a copy and shuffle the elements randomly. axis : int, optional The axis which `x` is shuffled along. Default is 0. Returns ------- out : ndarray Permuted sequence or array range. Examples -------- >>> rng = np.random.default_rng() >>> rng.permutation(10) array([1, 7, 4, 3, 0, 9, 2, 5, 8, 6]) # random >>> rng.permutation([1, 4, 9, 12, 15]) array([15, 1, 9, 4, 12]) # random >>> arr = np.arange(9).reshape((3, 3)) >>> rng.permutation(arr) array([[6, 7, 8], # random [0, 1, 2], [3, 4, 5]]) >>> rng.permutation("abc") Traceback (most recent call last): ... numpy.exceptions.AxisError: axis 0 is out of bounds for array of dimension 0 >>> arr = np.arange(9).reshape((3, 3)) >>> rng.permutation(arr, axis=1) array([[0, 2, 1], # random [3, 5, 4], [6, 8, 7]]) shuffle(x, axis=0) Modify an array or sequence in-place by shuffling its contents. The order of sub-arrays is changed but their contents remains the same. Parameters ---------- x : ndarray or MutableSequence The array, list or mutable sequence to be shuffled. axis : int, optional The axis which `x` is shuffled along. Default is 0. It is only supported on `ndarray` objects. Returns ------- None See Also -------- permuted permutation Notes ----- An important distinction between methods ``shuffle`` and ``permuted`` is how they both treat the ``axis`` parameter which can be found at :ref:`generator-handling-axis-parameter`. Examples -------- >>> rng = np.random.default_rng() >>> arr = np.arange(10) >>> arr array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9]) >>> rng.shuffle(arr) >>> arr array([2, 0, 7, 5, 1, 4, 8, 9, 3, 6]) # random >>> arr = np.arange(9).reshape((3, 3)) >>> arr array([[0, 1, 2], [3, 4, 5], [6, 7, 8]]) >>> rng.shuffle(arr) >>> arr array([[3, 4, 5], # random [6, 7, 8], [0, 1, 2]]) >>> arr = np.arange(9).reshape((3, 3)) >>> arr array([[0, 1, 2], [3, 4, 5], [6, 7, 8]]) >>> rng.shuffle(arr, axis=1) >>> arr array([[2, 0, 1], # random [5, 3, 4], [8, 6, 7]]) permuted(x, axis=None, out=None) Randomly permute `x` along axis `axis`. Unlike `shuffle`, each slice along the given axis is shuffled independently of the others. Parameters ---------- x : array_like, at least one-dimensional Array to be shuffled. axis : int, optional Slices of `x` in this axis are shuffled. Each slice is shuffled independently of the others. If `axis` is None, the flattened array is shuffled. out : ndarray, optional If given, this is the destination of the shuffled array. If `out` is None, a shuffled copy of the array is returned. Returns ------- ndarray If `out` is None, a shuffled copy of `x` is returned. Otherwise, the shuffled array is stored in `out`, and `out` is returned See Also -------- shuffle permutation Notes ----- An important distinction between methods ``shuffle`` and ``permuted`` is how they both treat the ``axis`` parameter which can be found at :ref:`generator-handling-axis-parameter`. Examples -------- Create a `numpy.random.Generator` instance: >>> rng = np.random.default_rng() Create a test array: >>> x = np.arange(24).reshape(3, 8) >>> x array([[ 0, 1, 2, 3, 4, 5, 6, 7], [ 8, 9, 10, 11, 12, 13, 14, 15], [16, 17, 18, 19, 20, 21, 22, 23]]) Shuffle the rows of `x`: >>> y = rng.permuted(x, axis=1) >>> y array([[ 4, 3, 6, 7, 1, 2, 5, 0], # random [15, 10, 14, 9, 12, 11, 8, 13], [17, 16, 20, 21, 18, 22, 23, 19]]) `x` has not been modified: >>> x array([[ 0, 1, 2, 3, 4, 5, 6, 7], [ 8, 9, 10, 11, 12, 13, 14, 15], [16, 17, 18, 19, 20, 21, 22, 23]]) To shuffle the rows of `x` in-place, pass `x` as the `out` parameter: >>> y = rng.permuted(x, axis=1, out=x) >>> x array([[ 3, 0, 4, 7, 1, 6, 2, 5], # random [ 8, 14, 13, 9, 12, 11, 15, 10], [17, 18, 16, 22, 19, 23, 20, 21]]) Note that when the ``out`` parameter is given, the return value is ``out``: >>> y is x True dirichlet(alpha, size=None) Draw samples from the Dirichlet distribution. Draw `size` samples of dimension k from a Dirichlet distribution. A Dirichlet-distributed random variable can be seen as a multivariate generalization of a Beta distribution. The Dirichlet distribution is a conjugate prior of a multinomial distribution in Bayesian inference. Parameters ---------- alpha : sequence of floats, length k Parameter of the distribution (length ``k`` for sample of length ``k``). size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n)``, then ``m * n * k`` samples are drawn. Default is None, in which case a vector of length ``k`` is returned. Returns ------- samples : ndarray, The drawn samples, of shape ``(size, k)``. Raises ------ ValueError If any value in ``alpha`` is less than zero Notes ----- The Dirichlet distribution is a distribution over vectors :math:`x` that fulfil the conditions :math:`x_i>0` and :math:`\sum_{i=1}^k x_i = 1`. The probability density function :math:`p` of a Dirichlet-distributed random vector :math:`X` is proportional to .. math:: p(x) \propto \prod_{i=1}^{k}{x^{\alpha_i-1}_i}, where :math:`\alpha` is a vector containing the positive concentration parameters. The method uses the following property for computation: let :math:`Y` be a random vector which has components that follow a standard gamma distribution, then :math:`X = \frac{1}{\sum_{i=1}^k{Y_i}} Y` is Dirichlet-distributed References ---------- .. [1] David McKay, "Information Theory, Inference and Learning Algorithms," chapter 23, https://www.inference.org.uk/mackay/itila/ .. [2] Wikipedia, "Dirichlet distribution", https://en.wikipedia.org/wiki/Dirichlet_distribution Examples -------- Taking an example cited in Wikipedia, this distribution can be used if one wanted to cut strings (each of initial length 1.0) into K pieces with different lengths, where each piece had, on average, a designated average length, but allowing some variation in the relative sizes of the pieces. >>> rng = np.random.default_rng() >>> s = rng.dirichlet((10, 5, 3), 20).transpose() >>> import matplotlib.pyplot as plt >>> plt.barh(range(20), s[0]) >>> plt.barh(range(20), s[1], left=s[0], color='g') >>> plt.barh(range(20), s[2], left=s[0]+s[1], color='r') >>> plt.title("Lengths of Strings") multivariate_hypergeometric(colors, nsample, size=None, method='marginals') Generate variates from a multivariate hypergeometric distribution. The multivariate hypergeometric distribution is a generalization of the hypergeometric distribution. Choose ``nsample`` items at random without replacement from a collection with ``N`` distinct types. ``N`` is the length of ``colors``, and the values in ``colors`` are the number of occurrences of that type in the collection. The total number of items in the collection is ``sum(colors)``. Each random variate generated by this function is a vector of length ``N`` holding the counts of the different types that occurred in the ``nsample`` items. The name ``colors`` comes from a common description of the distribution: it is the probability distribution of the number of marbles of each color selected without replacement from an urn containing marbles of different colors; ``colors[i]`` is the number of marbles in the urn with color ``i``. Parameters ---------- colors : sequence of integers The number of each type of item in the collection from which a sample is drawn. The values in ``colors`` must be nonnegative. To avoid loss of precision in the algorithm, ``sum(colors)`` must be less than ``10**9`` when `method` is "marginals". nsample : int The number of items selected. ``nsample`` must not be greater than ``sum(colors)``. size : int or tuple of ints, optional The number of variates to generate, either an integer or a tuple holding the shape of the array of variates. If the given size is, e.g., ``(k, m)``, then ``k * m`` variates are drawn, where one variate is a vector of length ``len(colors)``, and the return value has shape ``(k, m, len(colors))``. If `size` is an integer, the output has shape ``(size, len(colors))``. Default is None, in which case a single variate is returned as an array with shape ``(len(colors),)``. method : string, optional Specify the algorithm that is used to generate the variates. Must be 'count' or 'marginals' (the default). See the Notes for a description of the methods. Returns ------- variates : ndarray Array of variates drawn from the multivariate hypergeometric distribution. See Also -------- hypergeometric : Draw samples from the (univariate) hypergeometric distribution. Notes ----- The two methods do not return the same sequence of variates. The "count" algorithm is roughly equivalent to the following numpy code:: choices = np.repeat(np.arange(len(colors)), colors) selection = np.random.choice(choices, nsample, replace=False) variate = np.bincount(selection, minlength=len(colors)) The "count" algorithm uses a temporary array of integers with length ``sum(colors)``. The "marginals" algorithm generates a variate by using repeated calls to the univariate hypergeometric sampler. It is roughly equivalent to:: variate = np.zeros(len(colors), dtype=np.int64) # `remaining` is the cumulative sum of `colors` from the last # element to the first; e.g. if `colors` is [3, 1, 5], then # `remaining` is [9, 6, 5]. remaining = np.cumsum(colors[::-1])[::-1] for i in range(len(colors)-1): if nsample < 1: break variate[i] = hypergeometric(colors[i], remaining[i+1], nsample) nsample -= variate[i] variate[-1] = nsample The default method is "marginals". For some cases (e.g. when `colors` contains relatively small integers), the "count" method can be significantly faster than the "marginals" method. If performance of the algorithm is important, test the two methods with typical inputs to decide which works best. Examples -------- >>> colors = [16, 8, 4] >>> seed = 4861946401452 >>> gen = np.random.Generator(np.random.PCG64(seed)) >>> gen.multivariate_hypergeometric(colors, 6) array([5, 0, 1]) >>> gen.multivariate_hypergeometric(colors, 6, size=3) array([[5, 0, 1], [2, 2, 2], [3, 3, 0]]) >>> gen.multivariate_hypergeometric(colors, 6, size=(2, 2)) array([[[3, 2, 1], [3, 2, 1]], [[4, 1, 1], [3, 2, 1]]]) multinomial(n, pvals, size=None) Draw samples from a multinomial distribution. The multinomial distribution is a multivariate generalization of the binomial distribution. Take an experiment with one of ``p`` possible outcomes. An example of such an experiment is throwing a dice, where the outcome can be 1 through 6. Each sample drawn from the distribution represents `n` such experiments. Its values, ``X_i = [X_0, X_1, ..., X_p]``, represent the number of times the outcome was ``i``. Parameters ---------- n : int or array-like of ints Number of experiments. pvals : array-like of floats Probabilities of each of the ``p`` different outcomes with shape ``(k0, k1, ..., kn, p)``. Each element ``pvals[i,j,...,:]`` must sum to 1 (however, the last element is always assumed to account for the remaining probability, as long as ``sum(pvals[..., :-1], axis=-1) <= 1.0``. Must have at least 1 dimension where pvals.shape[-1] > 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn each with ``p`` elements. Default is None where the output size is determined by the broadcast shape of ``n`` and all by the final dimension of ``pvals``, which is denoted as ``b=(b0, b1, ..., bq)``. If size is not None, then it must be compatible with the broadcast shape ``b``. Specifically, size must have ``q`` or more elements and size[-(q-j):] must equal ``bj``. Returns ------- out : ndarray The drawn samples, of shape size, if provided. When size is provided, the output shape is size + (p,) If not specified, the shape is determined by the broadcast shape of ``n`` and ``pvals``, ``(b0, b1, ..., bq)`` augmented with the dimension of the multinomial, ``p``, so that that output shape is ``(b0, b1, ..., bq, p)``. Each entry ``out[i,j,...,:]`` is a ``p``-dimensional value drawn from the distribution. Examples -------- Throw a dice 20 times: >>> rng = np.random.default_rng() >>> rng.multinomial(20, [1/6.]*6, size=1) array([[4, 1, 7, 5, 2, 1]]) # random It landed 4 times on 1, once on 2, etc. Now, throw the dice 20 times, and 20 times again: >>> rng.multinomial(20, [1/6.]*6, size=2) array([[3, 4, 3, 3, 4, 3], [2, 4, 3, 4, 0, 7]]) # random For the first run, we threw 3 times 1, 4 times 2, etc. For the second, we threw 2 times 1, 4 times 2, etc. Now, do one experiment throwing the dice 10 time, and 10 times again, and another throwing the dice 20 times, and 20 times again: >>> rng.multinomial([[10], [20]], [1/6.]*6, size=(2, 2)) array([[[2, 4, 0, 1, 2, 1], [1, 3, 0, 3, 1, 2]], [[1, 4, 4, 4, 4, 3], [3, 3, 2, 5, 5, 2]]]) # random The first array shows the outcomes of throwing the dice 10 times, and the second shows the outcomes from throwing the dice 20 times. A loaded die is more likely to land on number 6: >>> rng.multinomial(100, [1/7.]*5 + [2/7.]) array([11, 16, 14, 17, 16, 26]) # random Simulate 10 throws of a 4-sided die and 20 throws of a 6-sided die >>> rng.multinomial([10, 20],[[1/4]*4 + [0]*2, [1/6]*6]) array([[2, 1, 4, 3, 0, 0], [3, 3, 3, 6, 1, 4]], dtype=int64) # random Generate categorical random variates from two categories where the first has 3 outcomes and the second has 2. >>> rng.multinomial(1, [[.1, .5, .4 ], [.3, .7, .0]]) array([[0, 0, 1], [0, 1, 0]], dtype=int64) # random ``argmax(axis=-1)`` is then used to return the categories. >>> pvals = [[.1, .5, .4 ], [.3, .7, .0]] >>> rvs = rng.multinomial(1, pvals, size=(4,2)) >>> rvs.argmax(axis=-1) array([[0, 1], [2, 0], [2, 1], [2, 0]], dtype=int64) # random The same output dimension can be produced using broadcasting. >>> rvs = rng.multinomial([[1]] * 4, pvals) >>> rvs.argmax(axis=-1) array([[0, 1], [2, 0], [2, 1], [2, 0]], dtype=int64) # random The probability inputs should be normalized. As an implementation detail, the value of the last entry is ignored and assumed to take up any leftover probability mass, but this should not be relied on. A biased coin which has twice as much weight on one side as on the other should be sampled like so: >>> rng.multinomial(100, [1.0 / 3, 2.0 / 3]) # RIGHT array([38, 62]) # random not like: >>> rng.multinomial(100, [1.0, 2.0]) # WRONG Traceback (most recent call last): ValueError: pvals < 0, pvals > 1 or pvals contains NaNs multivariate_normal(mean, cov, size=None, check_valid='warn', tol=1e-8, *, method='svd') Draw random samples from a multivariate normal distribution. The multivariate normal, multinormal or Gaussian distribution is a generalization of the one-dimensional normal distribution to higher dimensions. Such a distribution is specified by its mean and covariance matrix. These parameters are analogous to the mean (average or "center") and variance (the squared standard deviation, or "width") of the one-dimensional normal distribution. Parameters ---------- mean : 1-D array_like, of length N Mean of the N-dimensional distribution. cov : 2-D array_like, of shape (N, N) Covariance matrix of the distribution. It must be symmetric and positive-semidefinite for proper sampling. size : int or tuple of ints, optional Given a shape of, for example, ``(m,n,k)``, ``m*n*k`` samples are generated, and packed in an `m`-by-`n`-by-`k` arrangement. Because each sample is `N`-dimensional, the output shape is ``(m,n,k,N)``. If no shape is specified, a single (`N`-D) sample is returned. check_valid : { 'warn', 'raise', 'ignore' }, optional Behavior when the covariance matrix is not positive semidefinite. tol : float, optional Tolerance when checking the singular values in covariance matrix. cov is cast to double before the check. method : { 'svd', 'eigh', 'cholesky'}, optional The cov input is used to compute a factor matrix A such that ``A @ A.T = cov``. This argument is used to select the method used to compute the factor matrix A. The default method 'svd' is the slowest, while 'cholesky' is the fastest but less robust than the slowest method. The method `eigh` uses eigen decomposition to compute A and is faster than svd but slower than cholesky. Returns ------- out : ndarray The drawn samples, of shape *size*, if that was provided. If not, the shape is ``(N,)``. In other words, each entry ``out[i,j,...,:]`` is an N-dimensional value drawn from the distribution. Notes ----- The mean is a coordinate in N-dimensional space, which represents the location where samples are most likely to be generated. This is analogous to the peak of the bell curve for the one-dimensional or univariate normal distribution. Covariance indicates the level to which two variables vary together. From the multivariate normal distribution, we draw N-dimensional samples, :math:`X = [x_1, x_2, ... x_N]`. The covariance matrix element :math:`C_{ij}` is the covariance of :math:`x_i` and :math:`x_j`. The element :math:`C_{ii}` is the variance of :math:`x_i` (i.e. its "spread"). Instead of specifying the full covariance matrix, popular approximations include: - Spherical covariance (`cov` is a multiple of the identity matrix) - Diagonal covariance (`cov` has non-negative elements, and only on the diagonal) This geometrical property can be seen in two dimensions by plotting generated data-points: >>> mean = [0, 0] >>> cov = [[1, 0], [0, 100]] # diagonal covariance Diagonal covariance means that points are oriented along x or y-axis: >>> import matplotlib.pyplot as plt >>> rng = np.random.default_rng() >>> x, y = rng.multivariate_normal(mean, cov, 5000).T >>> plt.plot(x, y, 'x') >>> plt.axis('equal') >>> plt.show() Note that the covariance matrix must be positive semidefinite (a.k.a. nonnegative-definite). Otherwise, the behavior of this method is undefined and backwards compatibility is not guaranteed. This function internally uses linear algebra routines, and thus results may not be identical (even up to precision) across architectures, OSes, or even builds. For example, this is likely if ``cov`` has multiple equal singular values and ``method`` is ``'svd'`` (default). In this case, ``method='cholesky'`` may be more robust. References ---------- .. [1] Papoulis, A., "Probability, Random Variables, and Stochastic Processes," 3rd ed., New York: McGraw-Hill, 1991. .. [2] Duda, R. O., Hart, P. E., and Stork, D. G., "Pattern Classification," 2nd ed., New York: Wiley, 2001. Examples -------- >>> mean = (1, 2) >>> cov = [[1, 0], [0, 1]] >>> rng = np.random.default_rng() >>> x = rng.multivariate_normal(mean, cov, (3, 3)) >>> x.shape (3, 3, 2) We can use a different method other than the default to factorize cov: >>> y = rng.multivariate_normal(mean, cov, (3, 3), method='cholesky') >>> y.shape (3, 3, 2) Here we generate 800 samples from the bivariate normal distribution with mean [0, 0] and covariance matrix [[6, -3], [-3, 3.5]]. The expected variances of the first and second components of the sample are 6 and 3.5, respectively, and the expected correlation coefficient is -3/sqrt(6*3.5) ≈ -0.65465. >>> cov = np.array([[6, -3], [-3, 3.5]]) >>> pts = rng.multivariate_normal([0, 0], cov, size=800) Check that the mean, covariance, and correlation coefficient of the sample are close to the expected values: >>> pts.mean(axis=0) array([ 0.0326911 , -0.01280782]) # may vary >>> np.cov(pts.T) array([[ 5.96202397, -2.85602287], [-2.85602287, 3.47613949]]) # may vary >>> np.corrcoef(pts.T)[0, 1] -0.6273591314603949 # may vary We can visualize this data with a scatter plot. The orientation of the point cloud illustrates the negative correlation of the components of this sample. >>> import matplotlib.pyplot as plt >>> plt.plot(pts[:, 0], pts[:, 1], '.', alpha=0.5) >>> plt.axis('equal') >>> plt.grid() >>> plt.show() logseries(p, size=None) Draw samples from a logarithmic series distribution. Samples are drawn from a log series distribution with specified shape parameter, 0 <= ``p`` < 1. Parameters ---------- p : float or array_like of floats Shape parameter for the distribution. Must be in the range [0, 1). size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``p`` is a scalar. Otherwise, ``np.array(p).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized logarithmic series distribution. See Also -------- scipy.stats.logser : probability density function, distribution or cumulative density function, etc. Notes ----- The probability mass function for the Log Series distribution is .. math:: P(k) = \frac{-p^k}{k \ln(1-p)}, where p = probability. The log series distribution is frequently used to represent species richness and occurrence, first proposed by Fisher, Corbet, and Williams in 1943 [2]. It may also be used to model the numbers of occupants seen in cars [3]. References ---------- .. [1] Buzas, Martin A.; Culver, Stephen J., Understanding regional species diversity through the log series distribution of occurrences: BIODIVERSITY RESEARCH Diversity & Distributions, Volume 5, Number 5, September 1999 , pp. 187-195(9). .. [2] Fisher, R.A,, A.S. Corbet, and C.B. Williams. 1943. The relation between the number of species and the number of individuals in a random sample of an animal population. Journal of Animal Ecology, 12:42-58. .. [3] D. J. Hand, F. Daly, D. Lunn, E. Ostrowski, A Handbook of Small Data Sets, CRC Press, 1994. .. [4] Wikipedia, "Logarithmic distribution", https://en.wikipedia.org/wiki/Logarithmic_distribution Examples -------- Draw samples from the distribution: >>> a = .6 >>> rng = np.random.default_rng() >>> s = rng.logseries(a, 10000) >>> import matplotlib.pyplot as plt >>> bins = np.arange(-.5, max(s) + .5 ) >>> count, bins, _ = plt.hist(s, bins=bins, label='Sample count') # plot against distribution >>> def logseries(k, p): ... return -p**k/(k*np.log(1-p)) >>> centres = np.arange(1, max(s) + 1) >>> plt.plot(centres, logseries(centres, a) * s.size, 'r', label='logseries PMF') >>> plt.legend() >>> plt.show() hypergeometric(ngood, nbad, nsample, size=None) Draw samples from a Hypergeometric distribution. Samples are drawn from a hypergeometric distribution with specified parameters, `ngood` (ways to make a good selection), `nbad` (ways to make a bad selection), and `nsample` (number of items sampled, which is less than or equal to the sum ``ngood + nbad``). Parameters ---------- ngood : int or array_like of ints Number of ways to make a good selection. Must be nonnegative and less than 10**9. nbad : int or array_like of ints Number of ways to make a bad selection. Must be nonnegative and less than 10**9. nsample : int or array_like of ints Number of items sampled. Must be nonnegative and less than ``ngood + nbad``. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if `ngood`, `nbad`, and `nsample` are all scalars. Otherwise, ``np.broadcast(ngood, nbad, nsample).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized hypergeometric distribution. Each sample is the number of good items within a randomly selected subset of size `nsample` taken from a set of `ngood` good items and `nbad` bad items. See Also -------- multivariate_hypergeometric : Draw samples from the multivariate hypergeometric distribution. scipy.stats.hypergeom : probability density function, distribution or cumulative density function, etc. Notes ----- The probability mass function (PMF) for the Hypergeometric distribution is .. math:: P(x) = \frac{\binom{g}{x}\binom{b}{n-x}}{\binom{g+b}{n}}, where :math:`0 \le x \le n` and :math:`n-b \le x \le g` for P(x) the probability of ``x`` good results in the drawn sample, g = `ngood`, b = `nbad`, and n = `nsample`. Consider an urn with black and white marbles in it, `ngood` of them are black and `nbad` are white. If you draw `nsample` balls without replacement, then the hypergeometric distribution describes the distribution of black balls in the drawn sample. Note that this distribution is very similar to the binomial distribution, except that in this case, samples are drawn without replacement, whereas in the Binomial case samples are drawn with replacement (or the sample space is infinite). As the sample space becomes large, this distribution approaches the binomial. The arguments `ngood` and `nbad` each must be less than `10**9`. For extremely large arguments, the algorithm that is used to compute the samples [4]_ breaks down because of loss of precision in floating point calculations. For such large values, if `nsample` is not also large, the distribution can be approximated with the binomial distribution, `binomial(n=nsample, p=ngood/(ngood + nbad))`. References ---------- .. [1] Lentner, Marvin, "Elementary Applied Statistics", Bogden and Quigley, 1972. .. [2] Weisstein, Eric W. "Hypergeometric Distribution." From MathWorld--A Wolfram Web Resource. https://mathworld.wolfram.com/HypergeometricDistribution.html .. [3] Wikipedia, "Hypergeometric distribution", https://en.wikipedia.org/wiki/Hypergeometric_distribution .. [4] Stadlober, Ernst, "The ratio of uniforms approach for generating discrete random variates", Journal of Computational and Applied Mathematics, 31, pp. 181-189 (1990). Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> ngood, nbad, nsamp = 100, 2, 10 # number of good, number of bad, and number of samples >>> s = rng.hypergeometric(ngood, nbad, nsamp, 1000) >>> from matplotlib.pyplot import hist >>> hist(s) # note that it is very unlikely to grab both bad items Suppose you have an urn with 15 white and 15 black marbles. If you pull 15 marbles at random, how likely is it that 12 or more of them are one color? >>> s = rng.hypergeometric(15, 15, 15, 100000) >>> sum(s>=12)/100000. + sum(s<=3)/100000. # answer = 0.003 ... pretty unlikely! geometric(p, size=None) Draw samples from the geometric distribution. Bernoulli trials are experiments with one of two outcomes: success or failure (an example of such an experiment is flipping a coin). The geometric distribution models the number of trials that must be run in order to achieve success. It is therefore supported on the positive integers, ``k = 1, 2, ...``. The probability mass function of the geometric distribution is .. math:: f(k) = (1 - p)^{k - 1} p where `p` is the probability of success of an individual trial. Parameters ---------- p : float or array_like of floats The probability of success of an individual trial. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``p`` is a scalar. Otherwise, ``np.array(p).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized geometric distribution. References ---------- .. [1] Wikipedia, "Geometric distribution", https://en.wikipedia.org/wiki/Geometric_distribution Examples -------- Draw 10,000 values from the geometric distribution, with the probability of an individual success equal to ``p = 0.35``: >>> p, size = 0.35, 10000 >>> rng = np.random.default_rng() >>> sample = rng.geometric(p=p, size=size) What proportion of trials succeeded after a single run? >>> (sample == 1).sum()/size 0.34889999999999999 # may vary The geometric distribution with ``p=0.35`` looks as follows: >>> import matplotlib.pyplot as plt >>> count, bins, _ = plt.hist(sample, bins=30, density=True) >>> plt.plot(bins, (1-p)**(bins-1)*p) >>> plt.xlim([0, 25]) >>> plt.show() zipf(a, size=None) Draw samples from a Zipf distribution. Samples are drawn from a Zipf distribution with specified parameter `a` > 1. The Zipf distribution (also known as the zeta distribution) is a discrete probability distribution that satisfies Zipf's law: the frequency of an item is inversely proportional to its rank in a frequency table. Parameters ---------- a : float or array_like of floats Distribution parameter. Must be greater than 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``a`` is a scalar. Otherwise, ``np.array(a).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Zipf distribution. See Also -------- scipy.stats.zipf : probability density function, distribution, or cumulative density function, etc. Notes ----- The probability mass function (PMF) for the Zipf distribution is .. math:: p(k) = \frac{k^{-a}}{\zeta(a)}, for integers :math:`k \geq 1`, where :math:`\zeta` is the Riemann Zeta function. It is named for the American linguist George Kingsley Zipf, who noted that the frequency of any word in a sample of a language is inversely proportional to its rank in the frequency table. References ---------- .. [1] Zipf, G. K., "Selected Studies of the Principle of Relative Frequency in Language," Cambridge, MA: Harvard Univ. Press, 1932. Examples -------- Draw samples from the distribution: >>> a = 4.0 >>> n = 20000 >>> rng = np.random.default_rng() >>> s = rng.zipf(a, size=n) Display the histogram of the samples, along with the expected histogram based on the probability density function: >>> import matplotlib.pyplot as plt >>> from scipy.special import zeta # doctest: +SKIP `bincount` provides a fast histogram for small integers. >>> count = np.bincount(s) >>> k = np.arange(1, s.max() + 1) >>> plt.bar(k, count[1:], alpha=0.5, label='sample count') >>> plt.plot(k, n*(k**-a)/zeta(a), 'k.-', alpha=0.5, ... label='expected count') # doctest: +SKIP >>> plt.semilogy() >>> plt.grid(alpha=0.4) >>> plt.legend() >>> plt.title(f'Zipf sample, a={a}, size={n}') >>> plt.show() poisson(lam=1.0, size=None) Draw samples from a Poisson distribution. The Poisson distribution is the limit of the binomial distribution for large N. Parameters ---------- lam : float or array_like of floats Expected number of events occurring in a fixed-time interval, must be >= 0. A sequence must be broadcastable over the requested size. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``lam`` is a scalar. Otherwise, ``np.array(lam).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Poisson distribution. Notes ----- The probability mass function (PMF) of Poisson distribution is .. math:: f(k; \lambda)=\frac{\lambda^k e^{-\lambda}}{k!} For events with an expected separation :math:`\lambda` the Poisson distribution :math:`f(k; \lambda)` describes the probability of :math:`k` events occurring within the observed interval :math:`\lambda`. Because the output is limited to the range of the C int64 type, a ValueError is raised when `lam` is within 10 sigma of the maximum representable value. References ---------- .. [1] Weisstein, Eric W. "Poisson Distribution." From MathWorld--A Wolfram Web Resource. https://mathworld.wolfram.com/PoissonDistribution.html .. [2] Wikipedia, "Poisson distribution", https://en.wikipedia.org/wiki/Poisson_distribution Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> lam, size = 5, 10000 >>> s = rng.poisson(lam=lam, size=size) Verify the mean and variance, which should be approximately ``lam``: >>> s.mean(), s.var() (4.9917 5.1088311) # may vary Display the histogram and probability mass function: >>> import matplotlib.pyplot as plt >>> from scipy import stats >>> x = np.arange(0, 21) >>> pmf = stats.poisson.pmf(x, mu=lam) >>> plt.hist(s, bins=x, density=True, width=0.5) >>> plt.stem(x, pmf, 'C1-') >>> plt.show() Draw each 100 values for lambda 100 and 500: >>> s = rng.poisson(lam=(100., 500.), size=(100, 2)) negative_binomial(n, p, size=None) Draw samples from a negative binomial distribution. Samples are drawn from a negative binomial distribution with specified parameters, `n` successes and `p` probability of success where `n` is > 0 and `p` is in the interval (0, 1]. Parameters ---------- n : float or array_like of floats Parameter of the distribution, > 0. p : float or array_like of floats Parameter of the distribution. Must satisfy 0 < p <= 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``n`` and ``p`` are both scalars. Otherwise, ``np.broadcast(n, p).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized negative binomial distribution, where each sample is equal to N, the number of failures that occurred before a total of n successes was reached. Notes ----- The probability mass function of the negative binomial distribution is .. math:: P(N;n,p) = \frac{\Gamma(N+n)}{N!\Gamma(n)}p^{n}(1-p)^{N}, where :math:`n` is the number of successes, :math:`p` is the probability of success, :math:`N+n` is the number of trials, and :math:`\Gamma` is the gamma function. When :math:`n` is an integer, :math:`\frac{\Gamma(N+n)}{N!\Gamma(n)} = \binom{N+n-1}{N}`, which is the more common form of this term in the pmf. The negative binomial distribution gives the probability of N failures given n successes, with a success on the last trial. If one throws a die repeatedly until the third time a "1" appears, then the probability distribution of the number of non-"1"s that appear before the third "1" is a negative binomial distribution. Because this method internally calls ``Generator.poisson`` with an intermediate random value, a ValueError is raised when the choice of :math:`n` and :math:`p` would result in the mean + 10 sigma of the sampled intermediate distribution exceeding the max acceptable value of the ``Generator.poisson`` method. This happens when :math:`p` is too low (a lot of failures happen for every success) and :math:`n` is too big ( a lot of successes are allowed). Therefore, the :math:`n` and :math:`p` values must satisfy the constraint: .. math:: n\frac{1-p}{p}+10n\sqrt{n}\frac{1-p}{p}<2^{63}-1-10\sqrt{2^{63}-1}, Where the left side of the equation is the derived mean + 10 sigma of a sample from the gamma distribution internally used as the :math:`lam` parameter of a poisson sample, and the right side of the equation is the constraint for maximum value of :math:`lam` in ``Generator.poisson``. References ---------- .. [1] Weisstein, Eric W. "Negative Binomial Distribution." From MathWorld--A Wolfram Web Resource. https://mathworld.wolfram.com/NegativeBinomialDistribution.html .. [2] Wikipedia, "Negative binomial distribution", https://en.wikipedia.org/wiki/Negative_binomial_distribution Examples -------- Draw samples from the distribution: A real world example. A company drills wild-cat oil exploration wells, each with an estimated probability of success of 0.1. What is the probability of having one success for each successive well, that is what is the probability of a single success after drilling 5 wells, after 6 wells, etc.? >>> rng = np.random.default_rng() >>> s = rng.negative_binomial(1, 0.1, 100000) >>> for i in range(1, 11): # doctest: +SKIP ... probability = sum(s>> rng = np.random.default_rng() >>> n, p, size = 10, .5, 10000 >>> s = rng.binomial(n, p, 10000) Assume a company drills 9 wild-cat oil exploration wells, each with an estimated probability of success of ``p=0.1``. All nine wells fail. What is the probability of that happening? Over ``size = 20,000`` trials the probability of this happening is on average: >>> n, p, size = 9, 0.1, 20000 >>> np.sum(rng.binomial(n=n, p=p, size=size) == 0)/size 0.39015 # may vary The following can be used to visualize a sample with ``n=100``, ``p=0.4`` and the corresponding probability density function: >>> import matplotlib.pyplot as plt >>> from scipy.stats import binom >>> n, p, size = 100, 0.4, 10000 >>> sample = rng.binomial(n, p, size=size) >>> count, bins, _ = plt.hist(sample, 30, density=True) >>> x = np.arange(n) >>> y = binom.pmf(x, n, p) >>> plt.plot(x, y, linewidth=2, color='r') triangular(left, mode, right, size=None) Draw samples from the triangular distribution over the interval ``[left, right]``. The triangular distribution is a continuous probability distribution with lower limit left, peak at mode, and upper limit right. Unlike the other distributions, these parameters directly define the shape of the pdf. Parameters ---------- left : float or array_like of floats Lower limit. mode : float or array_like of floats The value where the peak of the distribution occurs. The value must fulfill the condition ``left <= mode <= right``. right : float or array_like of floats Upper limit, must be larger than `left`. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``left``, ``mode``, and ``right`` are all scalars. Otherwise, ``np.broadcast(left, mode, right).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized triangular distribution. Notes ----- The probability density function for the triangular distribution is .. math:: P(x;l, m, r) = \begin{cases} \frac{2(x-l)}{(r-l)(m-l)}& \text{for $l \leq x \leq m$},\\ \frac{2(r-x)}{(r-l)(r-m)}& \text{for $m \leq x \leq r$},\\ 0& \text{otherwise}. \end{cases} The triangular distribution is often used in ill-defined problems where the underlying distribution is not known, but some knowledge of the limits and mode exists. Often it is used in simulations. References ---------- .. [1] Wikipedia, "Triangular distribution" https://en.wikipedia.org/wiki/Triangular_distribution Examples -------- Draw values from the distribution and plot the histogram: >>> import matplotlib.pyplot as plt >>> rng = np.random.default_rng() >>> h = plt.hist(rng.triangular(-3, 0, 8, 100000), bins=200, ... density=True) >>> plt.show() wald(mean, scale, size=None) Draw samples from a Wald, or inverse Gaussian, distribution. As the scale approaches infinity, the distribution becomes more like a Gaussian. Some references claim that the Wald is an inverse Gaussian with mean equal to 1, but this is by no means universal. The inverse Gaussian distribution was first studied in relationship to Brownian motion. In 1956 M.C.K. Tweedie used the name inverse Gaussian because there is an inverse relationship between the time to cover a unit distance and distance covered in unit time. Parameters ---------- mean : float or array_like of floats Distribution mean, must be > 0. scale : float or array_like of floats Scale parameter, must be > 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``mean`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(mean, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Wald distribution. Notes ----- The probability density function for the Wald distribution is .. math:: P(x;mean,scale) = \sqrt{\frac{scale}{2\pi x^3}}e^ \frac{-scale(x-mean)^2}{2\cdotp mean^2x} As noted above the inverse Gaussian distribution first arise from attempts to model Brownian motion. It is also a competitor to the Weibull for use in reliability modeling and modeling stock returns and interest rate processes. References ---------- .. [1] Brighton Webs Ltd., Wald Distribution, https://web.archive.org/web/20090423014010/http://www.brighton-webs.co.uk:80/distributions/wald.asp .. [2] Chhikara, Raj S., and Folks, J. Leroy, "The Inverse Gaussian Distribution: Theory : Methodology, and Applications", CRC Press, 1988. .. [3] Wikipedia, "Inverse Gaussian distribution" https://en.wikipedia.org/wiki/Inverse_Gaussian_distribution Examples -------- Draw values from the distribution and plot the histogram: >>> import matplotlib.pyplot as plt >>> rng = np.random.default_rng() >>> h = plt.hist(rng.wald(3, 2, 100000), bins=200, density=True) >>> plt.show() rayleigh(scale=1.0, size=None) Draw samples from a Rayleigh distribution. The :math:`\chi` and Weibull distributions are generalizations of the Rayleigh. Parameters ---------- scale : float or array_like of floats, optional Scale, also equals the mode. Must be non-negative. Default is 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``scale`` is a scalar. Otherwise, ``np.array(scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Rayleigh distribution. Notes ----- The probability density function for the Rayleigh distribution is .. math:: P(x;scale) = \frac{x}{scale^2}e^{\frac{-x^2}{2 \cdotp scale^2}} The Rayleigh distribution would arise, for example, if the East and North components of the wind velocity had identical zero-mean Gaussian distributions. Then the wind speed would have a Rayleigh distribution. References ---------- .. [1] Brighton Webs Ltd., "Rayleigh Distribution," https://web.archive.org/web/20090514091424/http://brighton-webs.co.uk:80/distributions/rayleigh.asp .. [2] Wikipedia, "Rayleigh distribution" https://en.wikipedia.org/wiki/Rayleigh_distribution Examples -------- Draw values from the distribution and plot the histogram >>> from matplotlib.pyplot import hist >>> rng = np.random.default_rng() >>> values = hist(rng.rayleigh(3, 100000), bins=200, density=True) Wave heights tend to follow a Rayleigh distribution. If the mean wave height is 1 meter, what fraction of waves are likely to be larger than 3 meters? >>> meanvalue = 1 >>> modevalue = np.sqrt(2 / np.pi) * meanvalue >>> s = rng.rayleigh(modevalue, 1000000) The percentage of waves larger than 3 meters is: >>> 100.*sum(s>3)/1000000. 0.087300000000000003 # random lognormal(mean=0.0, sigma=1.0, size=None) Draw samples from a log-normal distribution. Draw samples from a log-normal distribution with specified mean, standard deviation, and array shape. Note that the mean and standard deviation are not the values for the distribution itself, but of the underlying normal distribution it is derived from. Parameters ---------- mean : float or array_like of floats, optional Mean value of the underlying normal distribution. Default is 0. sigma : float or array_like of floats, optional Standard deviation of the underlying normal distribution. Must be non-negative. Default is 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``mean`` and ``sigma`` are both scalars. Otherwise, ``np.broadcast(mean, sigma).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized log-normal distribution. See Also -------- scipy.stats.lognorm : probability density function, distribution, cumulative density function, etc. Notes ----- A variable `x` has a log-normal distribution if `log(x)` is normally distributed. The probability density function for the log-normal distribution is: .. math:: p(x) = \frac{1}{\sigma x \sqrt{2\pi}} e^{(-\frac{(ln(x)-\mu)^2}{2\sigma^2})} where :math:`\mu` is the mean and :math:`\sigma` is the standard deviation of the normally distributed logarithm of the variable. A log-normal distribution results if a random variable is the *product* of a large number of independent, identically-distributed variables in the same way that a normal distribution results if the variable is the *sum* of a large number of independent, identically-distributed variables. References ---------- .. [1] Limpert, E., Stahel, W. A., and Abbt, M., "Log-normal Distributions across the Sciences: Keys and Clues," BioScience, Vol. 51, No. 5, May, 2001. https://stat.ethz.ch/~stahel/lognormal/bioscience.pdf .. [2] Reiss, R.D. and Thomas, M., "Statistical Analysis of Extreme Values," Basel: Birkhauser Verlag, 2001, pp. 31-32. Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> mu, sigma = 3., 1. # mean and standard deviation >>> s = rng.lognormal(mu, sigma, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, _ = plt.hist(s, 100, density=True, align='mid') >>> x = np.linspace(min(bins), max(bins), 10000) >>> pdf = (np.exp(-(np.log(x) - mu)**2 / (2 * sigma**2)) ... / (x * sigma * np.sqrt(2 * np.pi))) >>> plt.plot(x, pdf, linewidth=2, color='r') >>> plt.axis('tight') >>> plt.show() Demonstrate that taking the products of random samples from a uniform distribution can be fit well by a log-normal probability density function. >>> # Generate a thousand samples: each is the product of 100 random >>> # values, drawn from a normal distribution. >>> rng = rng >>> b = [] >>> for i in range(1000): ... a = 10. + rng.standard_normal(100) ... b.append(np.prod(a)) >>> b = np.array(b) / np.min(b) # scale values to be positive >>> count, bins, _ = plt.hist(b, 100, density=True, align='mid') >>> sigma = np.std(np.log(b)) >>> mu = np.mean(np.log(b)) >>> x = np.linspace(min(bins), max(bins), 10000) >>> pdf = (np.exp(-(np.log(x) - mu)**2 / (2 * sigma**2)) ... / (x * sigma * np.sqrt(2 * np.pi))) >>> plt.plot(x, pdf, color='r', linewidth=2) >>> plt.show() logistic(loc=0.0, scale=1.0, size=None) Draw samples from a logistic distribution. Samples are drawn from a logistic distribution with specified parameters, loc (location or mean, also median), and scale (>0). Parameters ---------- loc : float or array_like of floats, optional Parameter of the distribution. Default is 0. scale : float or array_like of floats, optional Parameter of the distribution. Must be non-negative. Default is 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``loc`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(loc, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized logistic distribution. See Also -------- scipy.stats.logistic : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Logistic distribution is .. math:: P(x) = \frac{e^{-(x-\mu)/s}}{s(1+e^{-(x-\mu)/s})^2}, where :math:`\mu` = location and :math:`s` = scale. The Logistic distribution is used in Extreme Value problems where it can act as a mixture of Gumbel distributions, in Epidemiology, and by the World Chess Federation (FIDE) where it is used in the Elo ranking system, assuming the performance of each player is a logistically distributed random variable. References ---------- .. [1] Reiss, R.-D. and Thomas M. (2001), "Statistical Analysis of Extreme Values, from Insurance, Finance, Hydrology and Other Fields," Birkhauser Verlag, Basel, pp 132-133. .. [2] Weisstein, Eric W. "Logistic Distribution." From MathWorld--A Wolfram Web Resource. https://mathworld.wolfram.com/LogisticDistribution.html .. [3] Wikipedia, "Logistic-distribution", https://en.wikipedia.org/wiki/Logistic_distribution Examples -------- Draw samples from the distribution: >>> loc, scale = 10, 1 >>> rng = np.random.default_rng() >>> s = rng.logistic(loc, scale, 10000) >>> import matplotlib.pyplot as plt >>> count, bins, _ = plt.hist(s, bins=50, label='Sampled data') # plot sampled data against the exact distribution >>> def logistic(x, loc, scale): ... return np.exp((loc-x)/scale)/(scale*(1+np.exp((loc-x)/scale))**2) >>> logistic_values = logistic(bins, loc, scale) >>> bin_spacing = np.mean(np.diff(bins)) >>> plt.plot(bins, logistic_values * bin_spacing * s.size, label='Logistic PDF') >>> plt.legend() >>> plt.show() gumbel(loc=0.0, scale=1.0, size=None) Draw samples from a Gumbel distribution. Draw samples from a Gumbel distribution with specified location and scale. For more information on the Gumbel distribution, see Notes and References below. Parameters ---------- loc : float or array_like of floats, optional The location of the mode of the distribution. Default is 0. scale : float or array_like of floats, optional The scale parameter of the distribution. Default is 1. Must be non- negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``loc`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(loc, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Gumbel distribution. See Also -------- scipy.stats.gumbel_l scipy.stats.gumbel_r scipy.stats.genextreme weibull Notes ----- The Gumbel (or Smallest Extreme Value (SEV) or the Smallest Extreme Value Type I) distribution is one of a class of Generalized Extreme Value (GEV) distributions used in modeling extreme value problems. The Gumbel is a special case of the Extreme Value Type I distribution for maximums from distributions with "exponential-like" tails. The probability density for the Gumbel distribution is .. math:: p(x) = \frac{e^{-(x - \mu)/ \beta}}{\beta} e^{ -e^{-(x - \mu)/ \beta}}, where :math:`\mu` is the mode, a location parameter, and :math:`\beta` is the scale parameter. The Gumbel (named for German mathematician Emil Julius Gumbel) was used very early in the hydrology literature, for modeling the occurrence of flood events. It is also used for modeling maximum wind speed and rainfall rates. It is a "fat-tailed" distribution - the probability of an event in the tail of the distribution is larger than if one used a Gaussian, hence the surprisingly frequent occurrence of 100-year floods. Floods were initially modeled as a Gaussian process, which underestimated the frequency of extreme events. It is one of a class of extreme value distributions, the Generalized Extreme Value (GEV) distributions, which also includes the Weibull and Frechet. The function has a mean of :math:`\mu + 0.57721\beta` and a variance of :math:`\frac{\pi^2}{6}\beta^2`. References ---------- .. [1] Gumbel, E. J., "Statistics of Extremes," New York: Columbia University Press, 1958. .. [2] Reiss, R.-D. and Thomas, M., "Statistical Analysis of Extreme Values from Insurance, Finance, Hydrology and Other Fields," Basel: Birkhauser Verlag, 2001. Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> mu, beta = 0, 0.1 # location and scale >>> s = rng.gumbel(mu, beta, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, _ = plt.hist(s, 30, density=True) >>> plt.plot(bins, (1/beta)*np.exp(-(bins - mu)/beta) ... * np.exp( -np.exp( -(bins - mu) /beta) ), ... linewidth=2, color='r') >>> plt.show() Show how an extreme value distribution can arise from a Gaussian process and compare to a Gaussian: >>> means = [] >>> maxima = [] >>> for i in range(0,1000) : ... a = rng.normal(mu, beta, 1000) ... means.append(a.mean()) ... maxima.append(a.max()) >>> count, bins, _ = plt.hist(maxima, 30, density=True) >>> beta = np.std(maxima) * np.sqrt(6) / np.pi >>> mu = np.mean(maxima) - 0.57721*beta >>> plt.plot(bins, (1/beta)*np.exp(-(bins - mu)/beta) ... * np.exp(-np.exp(-(bins - mu)/beta)), ... linewidth=2, color='r') >>> plt.plot(bins, 1/(beta * np.sqrt(2 * np.pi)) ... * np.exp(-(bins - mu)**2 / (2 * beta**2)), ... linewidth=2, color='g') >>> plt.show() laplace(loc=0.0, scale=1.0, size=None) Draw samples from the Laplace or double exponential distribution with specified location (or mean) and scale (decay). The Laplace distribution is similar to the Gaussian/normal distribution, but is sharper at the peak and has fatter tails. It represents the difference between two independent, identically distributed exponential random variables. Parameters ---------- loc : float or array_like of floats, optional The position, :math:`\mu`, of the distribution peak. Default is 0. scale : float or array_like of floats, optional :math:`\lambda`, the exponential decay. Default is 1. Must be non- negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``loc`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(loc, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Laplace distribution. Notes ----- It has the probability density function .. math:: f(x; \mu, \lambda) = \frac{1}{2\lambda} \exp\left(-\frac{|x - \mu|}{\lambda}\right). The first law of Laplace, from 1774, states that the frequency of an error can be expressed as an exponential function of the absolute magnitude of the error, which leads to the Laplace distribution. For many problems in economics and health sciences, this distribution seems to model the data better than the standard Gaussian distribution. References ---------- .. [1] Abramowitz, M. and Stegun, I. A. (Eds.). "Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables, 9th printing," New York: Dover, 1972. .. [2] Kotz, Samuel, et. al. "The Laplace Distribution and Generalizations, " Birkhauser, 2001. .. [3] Weisstein, Eric W. "Laplace Distribution." From MathWorld--A Wolfram Web Resource. https://mathworld.wolfram.com/LaplaceDistribution.html .. [4] Wikipedia, "Laplace distribution", https://en.wikipedia.org/wiki/Laplace_distribution Examples -------- Draw samples from the distribution >>> loc, scale = 0., 1. >>> rng = np.random.default_rng() >>> s = rng.laplace(loc, scale, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, _ = plt.hist(s, 30, density=True) >>> x = np.arange(-8., 8., .01) >>> pdf = np.exp(-abs(x-loc)/scale)/(2.*scale) >>> plt.plot(x, pdf) Plot Gaussian for comparison: >>> g = (1/(scale * np.sqrt(2 * np.pi)) * ... np.exp(-(x - loc)**2 / (2 * scale**2))) >>> plt.plot(x,g) power(a, size=None) Draws samples in [0, 1] from a power distribution with positive exponent a - 1. Also known as the power function distribution. Parameters ---------- a : float or array_like of floats Parameter of the distribution. Must be non-negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``a`` is a scalar. Otherwise, ``np.array(a).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized power distribution. Raises ------ ValueError If a <= 0. Notes ----- The probability density function is .. math:: P(x; a) = ax^{a-1}, 0 \le x \le 1, a>0. The power function distribution is just the inverse of the Pareto distribution. It may also be seen as a special case of the Beta distribution. It is used, for example, in modeling the over-reporting of insurance claims. References ---------- .. [1] Christian Kleiber, Samuel Kotz, "Statistical size distributions in economics and actuarial sciences", Wiley, 2003. .. [2] Heckert, N. A. and Filliben, James J. "NIST Handbook 148: Dataplot Reference Manual, Volume 2: Let Subcommands and Library Functions", National Institute of Standards and Technology Handbook Series, June 2003. https://www.itl.nist.gov/div898/software/dataplot/refman2/auxillar/powpdf.pdf Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> a = 5. # shape >>> samples = 1000 >>> s = rng.power(a, samples) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, _ = plt.hist(s, bins=30) >>> x = np.linspace(0, 1, 100) >>> y = a*x**(a-1.) >>> normed_y = samples*np.diff(bins)[0]*y >>> plt.plot(x, normed_y) >>> plt.show() Compare the power function distribution to the inverse of the Pareto. >>> from scipy import stats # doctest: +SKIP >>> rvs = rng.power(5, 1000000) >>> rvsp = rng.pareto(5, 1000000) >>> xx = np.linspace(0,1,100) >>> powpdf = stats.powerlaw.pdf(xx,5) # doctest: +SKIP >>> plt.figure() >>> plt.hist(rvs, bins=50, density=True) >>> plt.plot(xx,powpdf,'r-') # doctest: +SKIP >>> plt.title('power(5)') >>> plt.figure() >>> plt.hist(1./(1.+rvsp), bins=50, density=True) >>> plt.plot(xx,powpdf,'r-') # doctest: +SKIP >>> plt.title('inverse of 1 + Generator.pareto(5)') >>> plt.figure() >>> plt.hist(1./(1.+rvsp), bins=50, density=True) >>> plt.plot(xx,powpdf,'r-') # doctest: +SKIP >>> plt.title('inverse of stats.pareto(5)') weibull(a, size=None) Draw samples from a Weibull distribution. Draw samples from a 1-parameter Weibull distribution with the given shape parameter `a`. .. math:: X = (-ln(U))^{1/a} Here, U is drawn from the uniform distribution over (0,1]. The more common 2-parameter Weibull, including a scale parameter :math:`\lambda` is just :math:`X = \lambda(-ln(U))^{1/a}`. Parameters ---------- a : float or array_like of floats Shape parameter of the distribution. Must be nonnegative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``a`` is a scalar. Otherwise, ``np.array(a).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Weibull distribution. See Also -------- scipy.stats.weibull_max scipy.stats.weibull_min scipy.stats.genextreme gumbel Notes ----- The Weibull (or Type III asymptotic extreme value distribution for smallest values, SEV Type III, or Rosin-Rammler distribution) is one of a class of Generalized Extreme Value (GEV) distributions used in modeling extreme value problems. This class includes the Gumbel and Frechet distributions. The probability density for the Weibull distribution is .. math:: p(x) = \frac{a} {\lambda}(\frac{x}{\lambda})^{a-1}e^{-(x/\lambda)^a}, where :math:`a` is the shape and :math:`\lambda` the scale. The function has its peak (the mode) at :math:`\lambda(\frac{a-1}{a})^{1/a}`. When ``a = 1``, the Weibull distribution reduces to the exponential distribution. References ---------- .. [1] Waloddi Weibull, Royal Technical University, Stockholm, 1939 "A Statistical Theory Of The Strength Of Materials", Ingeniorsvetenskapsakademiens Handlingar Nr 151, 1939, Generalstabens Litografiska Anstalts Forlag, Stockholm. .. [2] Waloddi Weibull, "A Statistical Distribution Function of Wide Applicability", Journal Of Applied Mechanics ASME Paper 1951. .. [3] Wikipedia, "Weibull distribution", https://en.wikipedia.org/wiki/Weibull_distribution Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> a = 5. # shape >>> s = rng.weibull(a, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> def weibull(x, n, a): ... return (a / n) * (x / n)**(a - 1) * np.exp(-(x / n)**a) >>> count, bins, _ = plt.hist(rng.weibull(5., 1000)) >>> x = np.linspace(0, 2, 1000) >>> bin_spacing = np.mean(np.diff(bins)) >>> plt.plot(x, weibull(x, 1., 5.) * bin_spacing * s.size, label='Weibull PDF') >>> plt.legend() >>> plt.show() pareto(a, size=None) Draw samples from a Pareto II (AKA Lomax) distribution with specified shape. Parameters ---------- a : float or array_like of floats Shape of the distribution. Must be positive. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``a`` is a scalar. Otherwise, ``np.array(a).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the Pareto II distribution. See Also -------- scipy.stats.pareto : Pareto I distribution scipy.stats.lomax : Lomax (Pareto II) distribution scipy.stats.genpareto : Generalized Pareto distribution Notes ----- The probability density for the Pareto II distribution is .. math:: p(x) = \frac{a}{{x+1}^{a+1}} , x \ge 0 where :math:`a > 0` is the shape. The Pareto II distribution is a shifted and scaled version of the Pareto I distribution, which can be found in `scipy.stats.pareto`. References ---------- .. [1] Francis Hunt and Paul Johnson, On the Pareto Distribution of Sourceforge projects. .. [2] Pareto, V. (1896). Course of Political Economy. Lausanne. .. [3] Reiss, R.D., Thomas, M.(2001), Statistical Analysis of Extreme Values, Birkhauser Verlag, Basel, pp 23-30. .. [4] Wikipedia, "Pareto distribution", https://en.wikipedia.org/wiki/Pareto_distribution Examples -------- Draw samples from the distribution: >>> a = 3. >>> rng = np.random.default_rng() >>> s = rng.pareto(a, 10000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> x = np.linspace(0, 3, 50) >>> pdf = a / (x+1)**(a+1) >>> plt.hist(s, bins=x, density=True, label='histogram') >>> plt.plot(x, pdf, linewidth=2, color='r', label='pdf') >>> plt.xlim(x.min(), x.max()) >>> plt.legend() >>> plt.show() vonmises(mu, kappa, size=None) Draw samples from a von Mises distribution. Samples are drawn from a von Mises distribution with specified mode (mu) and concentration (kappa), on the interval [-pi, pi]. The von Mises distribution (also known as the circular normal distribution) is a continuous probability distribution on the unit circle. It may be thought of as the circular analogue of the normal distribution. Parameters ---------- mu : float or array_like of floats Mode ("center") of the distribution. kappa : float or array_like of floats Concentration of the distribution, has to be >=0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``mu`` and ``kappa`` are both scalars. Otherwise, ``np.broadcast(mu, kappa).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized von Mises distribution. See Also -------- scipy.stats.vonmises : probability density function, distribution, or cumulative density function, etc. Notes ----- The probability density for the von Mises distribution is .. math:: p(x) = \frac{e^{\kappa cos(x-\mu)}}{2\pi I_0(\kappa)}, where :math:`\mu` is the mode and :math:`\kappa` the concentration, and :math:`I_0(\kappa)` is the modified Bessel function of order 0. The von Mises is named for Richard Edler von Mises, who was born in Austria-Hungary, in what is now the Ukraine. He fled to the United States in 1939 and became a professor at Harvard. He worked in probability theory, aerodynamics, fluid mechanics, and philosophy of science. References ---------- .. [1] Abramowitz, M. and Stegun, I. A. (Eds.). "Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables, 9th printing," New York: Dover, 1972. .. [2] von Mises, R., "Mathematical Theory of Probability and Statistics", New York: Academic Press, 1964. Examples -------- Draw samples from the distribution: >>> mu, kappa = 0.0, 4.0 # mean and concentration >>> rng = np.random.default_rng() >>> s = rng.vonmises(mu, kappa, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> from scipy.special import i0 # doctest: +SKIP >>> plt.hist(s, 50, density=True) >>> x = np.linspace(-np.pi, np.pi, num=51) >>> y = np.exp(kappa*np.cos(x-mu))/(2*np.pi*i0(kappa)) # doctest: +SKIP >>> plt.plot(x, y, linewidth=2, color='r') # doctest: +SKIP >>> plt.show() standard_t(df, size=None) Draw samples from a standard Student's t distribution with `df` degrees of freedom. A special case of the hyperbolic distribution. As `df` gets large, the result resembles that of the standard normal distribution (`standard_normal`). Parameters ---------- df : float or array_like of floats Degrees of freedom, must be > 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``df`` is a scalar. Otherwise, ``np.array(df).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized standard Student's t distribution. Notes ----- The probability density function for the t distribution is .. math:: P(x, df) = \frac{\Gamma(\frac{df+1}{2})}{\sqrt{\pi df} \Gamma(\frac{df}{2})}\Bigl( 1+\frac{x^2}{df} \Bigr)^{-(df+1)/2} The t test is based on an assumption that the data come from a Normal distribution. The t test provides a way to test whether the sample mean (that is the mean calculated from the data) is a good estimate of the true mean. The derivation of the t-distribution was first published in 1908 by William Gosset while working for the Guinness Brewery in Dublin. Due to proprietary issues, he had to publish under a pseudonym, and so he used the name Student. References ---------- .. [1] Dalgaard, Peter, "Introductory Statistics With R", Springer, 2002. .. [2] Wikipedia, "Student's t-distribution" https://en.wikipedia.org/wiki/Student's_t-distribution Examples -------- From Dalgaard page 83 [1]_, suppose the daily energy intake for 11 women in kilojoules (kJ) is: >>> intake = np.array([5260., 5470, 5640, 6180, 6390, 6515, 6805, 7515, \ ... 7515, 8230, 8770]) Does their energy intake deviate systematically from the recommended value of 7725 kJ? Our null hypothesis will be the absence of deviation, and the alternate hypothesis will be the presence of an effect that could be either positive or negative, hence making our test 2-tailed. Because we are estimating the mean and we have N=11 values in our sample, we have N-1=10 degrees of freedom. We set our significance level to 95% and compute the t statistic using the empirical mean and empirical standard deviation of our intake. We use a ddof of 1 to base the computation of our empirical standard deviation on an unbiased estimate of the variance (note: the final estimate is not unbiased due to the concave nature of the square root). >>> np.mean(intake) 6753.636363636364 >>> intake.std(ddof=1) 1142.1232221373727 >>> t = (np.mean(intake)-7725)/(intake.std(ddof=1)/np.sqrt(len(intake))) >>> t -2.8207540608310198 We draw 1000000 samples from Student's t distribution with the adequate degrees of freedom. >>> import matplotlib.pyplot as plt >>> rng = np.random.default_rng() >>> s = rng.standard_t(10, size=1000000) >>> h = plt.hist(s, bins=100, density=True) Does our t statistic land in one of the two critical regions found at both tails of the distribution? >>> np.sum(np.abs(t) < np.abs(s)) / float(len(s)) 0.018318 #random < 0.05, statistic is in critical region The probability value for this 2-tailed test is about 1.83%, which is lower than the 5% pre-determined significance threshold. Therefore, the probability of observing values as extreme as our intake conditionally on the null hypothesis being true is too low, and we reject the null hypothesis of no deviation. standard_cauchy(size=None) Draw samples from a standard Cauchy distribution with mode = 0. Also known as the Lorentz distribution. Parameters ---------- size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. Returns ------- samples : ndarray or scalar The drawn samples. Notes ----- The probability density function for the full Cauchy distribution is .. math:: P(x; x_0, \gamma) = \frac{1}{\pi \gamma \bigl[ 1+ (\frac{x-x_0}{\gamma})^2 \bigr] } and the Standard Cauchy distribution just sets :math:`x_0=0` and :math:`\gamma=1` The Cauchy distribution arises in the solution to the driven harmonic oscillator problem, and also describes spectral line broadening. It also describes the distribution of values at which a line tilted at a random angle will cut the x axis. When studying hypothesis tests that assume normality, seeing how the tests perform on data from a Cauchy distribution is a good indicator of their sensitivity to a heavy-tailed distribution, since the Cauchy looks very much like a Gaussian distribution, but with heavier tails. References ---------- .. [1] NIST/SEMATECH e-Handbook of Statistical Methods, "Cauchy Distribution", https://www.itl.nist.gov/div898/handbook/eda/section3/eda3663.htm .. [2] Weisstein, Eric W. "Cauchy Distribution." From MathWorld--A Wolfram Web Resource. https://mathworld.wolfram.com/CauchyDistribution.html .. [3] Wikipedia, "Cauchy distribution" https://en.wikipedia.org/wiki/Cauchy_distribution Examples -------- Draw samples and plot the distribution: >>> import matplotlib.pyplot as plt >>> rng = np.random.default_rng() >>> s = rng.standard_cauchy(1000000) >>> s = s[(s>-25) & (s<25)] # truncate distribution so it plots well >>> plt.hist(s, bins=100) >>> plt.show() noncentral_chisquare(df, nonc, size=None) Draw samples from a noncentral chi-square distribution. The noncentral :math:`\chi^2` distribution is a generalization of the :math:`\chi^2` distribution. Parameters ---------- df : float or array_like of floats Degrees of freedom, must be > 0. nonc : float or array_like of floats Non-centrality, must be non-negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``df`` and ``nonc`` are both scalars. Otherwise, ``np.broadcast(df, nonc).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized noncentral chi-square distribution. Notes ----- The probability density function for the noncentral Chi-square distribution is .. math:: P(x;df,nonc) = \sum^{\infty}_{i=0} \frac{e^{-nonc/2}(nonc/2)^{i}}{i!} P_{Y_{df+2i}}(x), where :math:`Y_{q}` is the Chi-square with q degrees of freedom. References ---------- .. [1] Wikipedia, "Noncentral chi-squared distribution" https://en.wikipedia.org/wiki/Noncentral_chi-squared_distribution Examples -------- Draw values from the distribution and plot the histogram >>> rng = np.random.default_rng() >>> import matplotlib.pyplot as plt >>> values = plt.hist(rng.noncentral_chisquare(3, 20, 100000), ... bins=200, density=True) >>> plt.show() Draw values from a noncentral chisquare with very small noncentrality, and compare to a chisquare. >>> plt.figure() >>> values = plt.hist(rng.noncentral_chisquare(3, .0000001, 100000), ... bins=np.arange(0., 25, .1), density=True) >>> values2 = plt.hist(rng.chisquare(3, 100000), ... bins=np.arange(0., 25, .1), density=True) >>> plt.plot(values[1][0:-1], values[0]-values2[0], 'ob') >>> plt.show() Demonstrate how large values of non-centrality lead to a more symmetric distribution. >>> plt.figure() >>> values = plt.hist(rng.noncentral_chisquare(3, 20, 100000), ... bins=200, density=True) >>> plt.show() chisquare(df, size=None) Draw samples from a chi-square distribution. When `df` independent random variables, each with standard normal distributions (mean 0, variance 1), are squared and summed, the resulting distribution is chi-square (see Notes). This distribution is often used in hypothesis testing. Parameters ---------- df : float or array_like of floats Number of degrees of freedom, must be > 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``df`` is a scalar. Otherwise, ``np.array(df).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized chi-square distribution. Raises ------ ValueError When `df` <= 0 or when an inappropriate `size` (e.g. ``size=-1``) is given. Notes ----- The variable obtained by summing the squares of `df` independent, standard normally distributed random variables: .. math:: Q = \sum_{i=1}^{\mathtt{df}} X^2_i is chi-square distributed, denoted .. math:: Q \sim \chi^2_k. The probability density function of the chi-squared distribution is .. math:: p(x) = \frac{(1/2)^{k/2}}{\Gamma(k/2)} x^{k/2 - 1} e^{-x/2}, where :math:`\Gamma` is the gamma function, .. math:: \Gamma(x) = \int_0^{-\infty} t^{x - 1} e^{-t} dt. References ---------- .. [1] NIST "Engineering Statistics Handbook" https://www.itl.nist.gov/div898/handbook/eda/section3/eda3666.htm Examples -------- >>> rng = np.random.default_rng() >>> rng.chisquare(2,4) array([ 1.89920014, 9.00867716, 3.13710533, 5.62318272]) # random The distribution of a chi-square random variable with 20 degrees of freedom looks as follows: >>> import matplotlib.pyplot as plt >>> import scipy.stats as stats >>> s = rng.chisquare(20, 10000) >>> count, bins, _ = plt.hist(s, 30, density=True) >>> x = np.linspace(0, 60, 1000) >>> plt.plot(x, stats.chi2.pdf(x, df=20)) >>> plt.xlim([0, 60]) >>> plt.show() noncentral_f(dfnum, dfden, nonc, size=None) Draw samples from the noncentral F distribution. Samples are drawn from an F distribution with specified parameters, `dfnum` (degrees of freedom in numerator) and `dfden` (degrees of freedom in denominator), where both parameters > 1. `nonc` is the non-centrality parameter. Parameters ---------- dfnum : float or array_like of floats Numerator degrees of freedom, must be > 0. dfden : float or array_like of floats Denominator degrees of freedom, must be > 0. nonc : float or array_like of floats Non-centrality parameter, the sum of the squares of the numerator means, must be >= 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``dfnum``, ``dfden``, and ``nonc`` are all scalars. Otherwise, ``np.broadcast(dfnum, dfden, nonc).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized noncentral Fisher distribution. Notes ----- When calculating the power of an experiment (power = probability of rejecting the null hypothesis when a specific alternative is true) the non-central F statistic becomes important. When the null hypothesis is true, the F statistic follows a central F distribution. When the null hypothesis is not true, then it follows a non-central F statistic. References ---------- .. [1] Weisstein, Eric W. "Noncentral F-Distribution." From MathWorld--A Wolfram Web Resource. https://mathworld.wolfram.com/NoncentralF-Distribution.html .. [2] Wikipedia, "Noncentral F-distribution", https://en.wikipedia.org/wiki/Noncentral_F-distribution Examples -------- In a study, testing for a specific alternative to the null hypothesis requires use of the Noncentral F distribution. We need to calculate the area in the tail of the distribution that exceeds the value of the F distribution for the null hypothesis. We'll plot the two probability distributions for comparison. >>> rng = np.random.default_rng() >>> dfnum = 3 # between group deg of freedom >>> dfden = 20 # within groups degrees of freedom >>> nonc = 3.0 >>> nc_vals = rng.noncentral_f(dfnum, dfden, nonc, 1000000) >>> NF = np.histogram(nc_vals, bins=50, density=True) >>> c_vals = rng.f(dfnum, dfden, 1000000) >>> F = np.histogram(c_vals, bins=50, density=True) >>> import matplotlib.pyplot as plt >>> plt.plot(F[1][1:], F[0]) >>> plt.plot(NF[1][1:], NF[0]) >>> plt.show() f(dfnum, dfden, size=None) Draw samples from an F distribution. Samples are drawn from an F distribution with specified parameters, `dfnum` (degrees of freedom in numerator) and `dfden` (degrees of freedom in denominator), where both parameters must be greater than zero. The random variate of the F distribution (also known as the Fisher distribution) is a continuous probability distribution that arises in ANOVA tests, and is the ratio of two chi-square variates. Parameters ---------- dfnum : float or array_like of floats Degrees of freedom in numerator, must be > 0. dfden : float or array_like of float Degrees of freedom in denominator, must be > 0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``dfnum`` and ``dfden`` are both scalars. Otherwise, ``np.broadcast(dfnum, dfden).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized Fisher distribution. See Also -------- scipy.stats.f : probability density function, distribution or cumulative density function, etc. Notes ----- The F statistic is used to compare in-group variances to between-group variances. Calculating the distribution depends on the sampling, and so it is a function of the respective degrees of freedom in the problem. The variable `dfnum` is the number of samples minus one, the between-groups degrees of freedom, while `dfden` is the within-groups degrees of freedom, the sum of the number of samples in each group minus the number of groups. References ---------- .. [1] Glantz, Stanton A. "Primer of Biostatistics.", McGraw-Hill, Fifth Edition, 2002. .. [2] Wikipedia, "F-distribution", https://en.wikipedia.org/wiki/F-distribution Examples -------- An example from Glantz[1], pp 47-40: Two groups, children of diabetics (25 people) and children from people without diabetes (25 controls). Fasting blood glucose was measured, case group had a mean value of 86.1, controls had a mean value of 82.2. Standard deviations were 2.09 and 2.49 respectively. Are these data consistent with the null hypothesis that the parents diabetic status does not affect their children's blood glucose levels? Calculating the F statistic from the data gives a value of 36.01. Draw samples from the distribution: >>> dfnum = 1. # between group degrees of freedom >>> dfden = 48. # within groups degrees of freedom >>> rng = np.random.default_rng() >>> s = rng.f(dfnum, dfden, 1000) The lower bound for the top 1% of the samples is : >>> np.sort(s)[-10] 7.61988120985 # random So there is about a 1% chance that the F statistic will exceed 7.62, the measured value is 36, so the null hypothesis is rejected at the 1% level. The corresponding probability density function for ``n = 20`` and ``m = 20`` is: >>> import matplotlib.pyplot as plt >>> from scipy import stats >>> dfnum, dfden, size = 20, 20, 10000 >>> s = rng.f(dfnum=dfnum, dfden=dfden, size=size) >>> bins, density, _ = plt.hist(s, 30, density=True) >>> x = np.linspace(0, 5, 1000) >>> plt.plot(x, stats.f.pdf(x, dfnum, dfden)) >>> plt.xlim([0, 5]) >>> plt.show() gamma(shape, scale=1.0, size=None) Draw samples from a Gamma distribution. Samples are drawn from a Gamma distribution with specified parameters, `shape` (sometimes designated "k") and `scale` (sometimes designated "theta"), where both parameters are > 0. Parameters ---------- shape : float or array_like of floats The shape of the gamma distribution. Must be non-negative. scale : float or array_like of floats, optional The scale of the gamma distribution. Must be non-negative. Default is equal to 1. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``shape`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(shape, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized gamma distribution. See Also -------- scipy.stats.gamma : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Gamma distribution is .. math:: p(x) = x^{k-1}\frac{e^{-x/\theta}}{\theta^k\Gamma(k)}, where :math:`k` is the shape and :math:`\theta` the scale, and :math:`\Gamma` is the Gamma function. The Gamma distribution is often used to model the times to failure of electronic components, and arises naturally in processes for which the waiting times between Poisson distributed events are relevant. References ---------- .. [1] Weisstein, Eric W. "Gamma Distribution." From MathWorld--A Wolfram Web Resource. https://mathworld.wolfram.com/GammaDistribution.html .. [2] Wikipedia, "Gamma distribution", https://en.wikipedia.org/wiki/Gamma_distribution Examples -------- Draw samples from the distribution: >>> shape, scale = 2., 2. # mean=4, std=2*sqrt(2) >>> rng = np.random.default_rng() >>> s = rng.gamma(shape, scale, 1000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> import scipy.special as sps # doctest: +SKIP >>> count, bins, _ = plt.hist(s, 50, density=True) >>> y = bins**(shape-1)*(np.exp(-bins/scale) / # doctest: +SKIP ... (sps.gamma(shape)*scale**shape)) >>> plt.plot(bins, y, linewidth=2, color='r') # doctest: +SKIP >>> plt.show() standard_gamma(shape, size=None, dtype=np.float64, out=None) Draw samples from a standard Gamma distribution. Samples are drawn from a Gamma distribution with specified parameters, shape (sometimes designated "k") and scale=1. Parameters ---------- shape : float or array_like of floats Parameter, must be non-negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``shape`` is a scalar. Otherwise, ``np.array(shape).size`` samples are drawn. dtype : dtype, optional Desired dtype of the result, only `float64` and `float32` are supported. Byteorder must be native. The default value is np.float64. out : ndarray, optional Alternative output array in which to place the result. If size is not None, it must have the same shape as the provided size and must match the type of the output values. Returns ------- out : ndarray or scalar Drawn samples from the parameterized standard gamma distribution. See Also -------- scipy.stats.gamma : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Gamma distribution is .. math:: p(x) = x^{k-1}\frac{e^{-x/\theta}}{\theta^k\Gamma(k)}, where :math:`k` is the shape and :math:`\theta` the scale, and :math:`\Gamma` is the Gamma function. The Gamma distribution is often used to model the times to failure of electronic components, and arises naturally in processes for which the waiting times between Poisson distributed events are relevant. References ---------- .. [1] Weisstein, Eric W. "Gamma Distribution." From MathWorld--A Wolfram Web Resource. https://mathworld.wolfram.com/GammaDistribution.html .. [2] Wikipedia, "Gamma distribution", https://en.wikipedia.org/wiki/Gamma_distribution Examples -------- Draw samples from the distribution: >>> shape, scale = 2., 1. # mean and width >>> rng = np.random.default_rng() >>> s = rng.standard_gamma(shape, 1000000) Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> import scipy.special as sps # doctest: +SKIP >>> count, bins, _ = plt.hist(s, 50, density=True) >>> y = bins**(shape-1) * ((np.exp(-bins/scale))/ # doctest: +SKIP ... (sps.gamma(shape) * scale**shape)) >>> plt.plot(bins, y, linewidth=2, color='r') # doctest: +SKIP >>> plt.show() normal(loc=0.0, scale=1.0, size=None) Draw random samples from a normal (Gaussian) distribution. The probability density function of the normal distribution, first derived by De Moivre and 200 years later by both Gauss and Laplace independently [2]_, is often called the bell curve because of its characteristic shape (see the example below). The normal distributions occurs often in nature. For example, it describes the commonly occurring distribution of samples influenced by a large number of tiny, random disturbances, each with its own unique distribution [2]_. Parameters ---------- loc : float or array_like of floats Mean ("centre") of the distribution. scale : float or array_like of floats Standard deviation (spread or "width") of the distribution. Must be non-negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``loc`` and ``scale`` are both scalars. Otherwise, ``np.broadcast(loc, scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized normal distribution. See Also -------- scipy.stats.norm : probability density function, distribution or cumulative density function, etc. Notes ----- The probability density for the Gaussian distribution is .. math:: p(x) = \frac{1}{\sqrt{ 2 \pi \sigma^2 }} e^{ - \frac{ (x - \mu)^2 } {2 \sigma^2} }, where :math:`\mu` is the mean and :math:`\sigma` the standard deviation. The square of the standard deviation, :math:`\sigma^2`, is called the variance. The function has its peak at the mean, and its "spread" increases with the standard deviation (the function reaches 0.607 times its maximum at :math:`x + \sigma` and :math:`x - \sigma` [2]_). This implies that :meth:`normal` is more likely to return samples lying close to the mean, rather than those far away. References ---------- .. [1] Wikipedia, "Normal distribution", https://en.wikipedia.org/wiki/Normal_distribution .. [2] P. R. Peebles Jr., "Central Limit Theorem" in "Probability, Random Variables and Random Signal Principles", 4th ed., 2001, pp. 51, 51, 125. Examples -------- Draw samples from the distribution: >>> mu, sigma = 0, 0.1 # mean and standard deviation >>> rng = np.random.default_rng() >>> s = rng.normal(mu, sigma, 1000) Verify the mean and the standard deviation: >>> abs(mu - np.mean(s)) 0.0 # may vary >>> abs(sigma - np.std(s, ddof=1)) 0.0 # may vary Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, _ = plt.hist(s, 30, density=True) >>> plt.plot(bins, 1/(sigma * np.sqrt(2 * np.pi)) * ... np.exp( - (bins - mu)**2 / (2 * sigma**2) ), ... linewidth=2, color='r') >>> plt.show() Two-by-four array of samples from the normal distribution with mean 3 and standard deviation 2.5: >>> rng = np.random.default_rng() >>> rng.normal(3, 2.5, size=(2, 4)) array([[-4.49401501, 4.00950034, -1.81814867, 7.29718677], # random [ 0.39924804, 4.68456316, 4.99394529, 4.84057254]]) # random standard_normal(size=None, dtype=np.float64, out=None) Draw samples from a standard Normal distribution (mean=0, stdev=1). Parameters ---------- size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. dtype : dtype, optional Desired dtype of the result, only `float64` and `float32` are supported. Byteorder must be native. The default value is np.float64. out : ndarray, optional Alternative output array in which to place the result. If size is not None, it must have the same shape as the provided size and must match the type of the output values. Returns ------- out : float or ndarray A floating-point array of shape ``size`` of drawn samples, or a single sample if ``size`` was not specified. See Also -------- normal : Equivalent function with additional ``loc`` and ``scale`` arguments for setting the mean and standard deviation. Notes ----- For random samples from the normal distribution with mean ``mu`` and standard deviation ``sigma``, use one of:: mu + sigma * rng.standard_normal(size=...) rng.normal(mu, sigma, size=...) Examples -------- >>> rng = np.random.default_rng() >>> rng.standard_normal() 2.1923875335537315 # random >>> s = rng.standard_normal(8000) >>> s array([ 0.6888893 , 0.78096262, -0.89086505, ..., 0.49876311, # random -0.38672696, -0.4685006 ]) # random >>> s.shape (8000,) >>> s = rng.standard_normal(size=(3, 4, 2)) >>> s.shape (3, 4, 2) Two-by-four array of samples from the normal distribution with mean 3 and standard deviation 2.5: >>> 3 + 2.5 * rng.standard_normal(size=(2, 4)) array([[-4.49401501, 4.00950034, -1.81814867, 7.29718677], # random [ 0.39924804, 4.68456316, 4.99394529, 4.84057254]]) # random uniform(low=0.0, high=1.0, size=None) Draw samples from a uniform distribution. Samples are uniformly distributed over the half-open interval ``[low, high)`` (includes low, but excludes high). In other words, any value within the given interval is equally likely to be drawn by `uniform`. Parameters ---------- low : float or array_like of floats, optional Lower boundary of the output interval. All values generated will be greater than or equal to low. The default value is 0. high : float or array_like of floats Upper boundary of the output interval. All values generated will be less than high. The high limit may be included in the returned array of floats due to floating-point rounding in the equation ``low + (high-low) * random_sample()``. high - low must be non-negative. The default value is 1.0. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``low`` and ``high`` are both scalars. Otherwise, ``np.broadcast(low, high).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized uniform distribution. See Also -------- integers : Discrete uniform distribution, yielding integers. random : Floats uniformly distributed over ``[0, 1)``. Notes ----- The probability density function of the uniform distribution is .. math:: p(x) = \frac{1}{b - a} anywhere within the interval ``[a, b)``, and zero elsewhere. When ``high`` == ``low``, values of ``low`` will be returned. Examples -------- Draw samples from the distribution: >>> rng = np.random.default_rng() >>> s = rng.uniform(-1,0,1000) All values are within the given interval: >>> np.all(s >= -1) True >>> np.all(s < 0) True Display the histogram of the samples, along with the probability density function: >>> import matplotlib.pyplot as plt >>> count, bins, _ = plt.hist(s, 15, density=True) >>> plt.plot(bins, np.ones_like(bins), linewidth=2, color='r') >>> plt.show() choice(a, size=None, replace=True, p=None, axis=0, shuffle=True) Generates a random sample from a given array Parameters ---------- a : {array_like, int} If an ndarray, a random sample is generated from its elements. If an int, the random sample is generated from np.arange(a). size : {int, tuple[int]}, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn from the 1-d `a`. If `a` has more than one dimension, the `size` shape will be inserted into the `axis` dimension, so the output ``ndim`` will be ``a.ndim - 1 + len(size)``. Default is None, in which case a single value is returned. replace : bool, optional Whether the sample is with or without replacement. Default is True, meaning that a value of ``a`` can be selected multiple times. p : 1-D array_like, optional The probabilities associated with each entry in a. If not given, the sample assumes a uniform distribution over all entries in ``a``. axis : int, optional The axis along which the selection is performed. The default, 0, selects by row. shuffle : bool, optional Whether the sample is shuffled when sampling without replacement. Default is True, False provides a speedup. Returns ------- samples : single item or ndarray The generated random samples Raises ------ ValueError If a is an int and less than zero, if p is not 1-dimensional, if a is array-like with a size 0, if p is not a vector of probabilities, if a and p have different lengths, or if replace=False and the sample size is greater than the population size. See Also -------- integers, shuffle, permutation Notes ----- Setting user-specified probabilities through ``p`` uses a more general but less efficient sampler than the default. The general sampler produces a different sample than the optimized sampler even if each element of ``p`` is 1 / len(a). ``p`` must sum to 1 when cast to ``float64``. To ensure this, you may wish to normalize using ``p = p / np.sum(p, dtype=float)``. When passing ``a`` as an integer type and ``size`` is not specified, the return type is a native Python ``int``. Examples -------- Generate a uniform random sample from np.arange(5) of size 3: >>> rng = np.random.default_rng() >>> rng.choice(5, 3) array([0, 3, 4]) # random >>> #This is equivalent to rng.integers(0,5,3) Generate a non-uniform random sample from np.arange(5) of size 3: >>> rng.choice(5, 3, p=[0.1, 0, 0.3, 0.6, 0]) array([3, 3, 0]) # random Generate a uniform random sample from np.arange(5) of size 3 without replacement: >>> rng.choice(5, 3, replace=False) array([3,1,0]) # random >>> #This is equivalent to rng.permutation(np.arange(5))[:3] Generate a uniform random sample from a 2-D array along the first axis (the default), without replacement: >>> rng.choice([[0, 1, 2], [3, 4, 5], [6, 7, 8]], 2, replace=False) array([[3, 4, 5], # random [0, 1, 2]]) Generate a non-uniform random sample from np.arange(5) of size 3 without replacement: >>> rng.choice(5, 3, replace=False, p=[0.1, 0, 0.3, 0.6, 0]) array([2, 3, 0]) # random Any of the above can be repeated with an arbitrary array-like instead of just integers. For instance: >>> aa_milne_arr = ['pooh', 'rabbit', 'piglet', 'Christopher'] >>> rng.choice(aa_milne_arr, 5, p=[0.5, 0.1, 0.1, 0.3]) array(['pooh', 'pooh', 'pooh', 'Christopher', 'piglet'], # random dtype='>> rng = np.random.default_rng() >>> rng.bytes(10) b'\xfeC\x9b\x86\x17\xf2\xa1\xafcp' # random integers(low, high=None, size=None, dtype=np.int64, endpoint=False) Return random integers from `low` (inclusive) to `high` (exclusive), or if endpoint=True, `low` (inclusive) to `high` (inclusive). Replaces `RandomState.randint` (with endpoint=False) and `RandomState.random_integers` (with endpoint=True) Return random integers from the "discrete uniform" distribution of the specified dtype. If `high` is None (the default), then results are from 0 to `low`. Parameters ---------- low : int or array-like of ints Lowest (signed) integers to be drawn from the distribution (unless ``high=None``, in which case this parameter is 0 and this value is used for `high`). high : int or array-like of ints, optional If provided, one above the largest (signed) integer to be drawn from the distribution (see above for behavior if ``high=None``). If array-like, must contain integer values size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. dtype : dtype, optional Desired dtype of the result. Byteorder must be native. The default value is np.int64. endpoint : bool, optional If true, sample from the interval [low, high] instead of the default [low, high) Defaults to False Returns ------- out : int or ndarray of ints `size`-shaped array of random integers from the appropriate distribution, or a single such random int if `size` not provided. Notes ----- When using broadcasting with uint64 dtypes, the maximum value (2**64) cannot be represented as a standard integer type. The high array (or low if high is None) must have object dtype, e.g., array([2**64]). Examples -------- >>> rng = np.random.default_rng() >>> rng.integers(2, size=10) array([1, 0, 0, 0, 1, 1, 0, 0, 1, 0]) # random >>> rng.integers(1, size=10) array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0]) Generate a 2 x 4 array of ints between 0 and 4, inclusive: >>> rng.integers(5, size=(2, 4)) array([[4, 0, 2, 1], [3, 2, 2, 0]]) # random Generate a 1 x 3 array with 3 different upper bounds >>> rng.integers(1, [3, 5, 10]) array([2, 2, 9]) # random Generate a 1 by 3 array with 3 different lower bounds >>> rng.integers([1, 5, 7], 10) array([9, 8, 7]) # random Generate a 2 by 4 array using broadcasting with dtype of uint8 >>> rng.integers([1, 3, 5, 7], [[10], [20]], dtype=np.uint8) array([[ 8, 6, 9, 7], [ 1, 16, 9, 12]], dtype=uint8) # random References ---------- .. [1] Daniel Lemire., "Fast Random Integer Generation in an Interval", ACM Transactions on Modeling and Computer Simulation 29 (1), 2019, https://arxiv.org/abs/1805.10941. standard_exponential(size=None, dtype=np.float64, method='zig', out=None) Draw samples from the standard exponential distribution. `standard_exponential` is identical to the exponential distribution with a scale parameter of 1. Parameters ---------- size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. dtype : dtype, optional Desired dtype of the result, only `float64` and `float32` are supported. Byteorder must be native. The default value is np.float64. method : str, optional Either 'inv' or 'zig'. 'inv' uses the default inverse CDF method. 'zig' uses the much faster Ziggurat method of Marsaglia and Tsang. out : ndarray, optional Alternative output array in which to place the result. If size is not None, it must have the same shape as the provided size and must match the type of the output values. Returns ------- out : float or ndarray Drawn samples. Examples -------- Output a 3x8000 array: >>> rng = np.random.default_rng() >>> n = rng.standard_exponential((3, 8000)) exponential(scale=1.0, size=None) Draw samples from an exponential distribution. Its probability density function is .. math:: f(x; \frac{1}{\beta}) = \frac{1}{\beta} \exp(-\frac{x}{\beta}), for ``x > 0`` and 0 elsewhere. :math:`\beta` is the scale parameter, which is the inverse of the rate parameter :math:`\lambda = 1/\beta`. The rate parameter is an alternative, widely used parameterization of the exponential distribution [3]_. The exponential distribution is a continuous analogue of the geometric distribution. It describes many common situations, such as the size of raindrops measured over many rainstorms [1]_, or the time between page requests to Wikipedia [2]_. Parameters ---------- scale : float or array_like of floats The scale parameter, :math:`\beta = 1/\lambda`. Must be non-negative. size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``scale`` is a scalar. Otherwise, ``np.array(scale).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized exponential distribution. Examples -------- Assume a company has 10000 customer support agents and the time between customer calls is exponentially distributed and that the average time between customer calls is 4 minutes. >>> scale, size = 4, 10000 >>> rng = np.random.default_rng() >>> time_between_calls = rng.exponential(scale=scale, size=size) What is the probability that a customer will call in the next 4 to 5 minutes? >>> x = ((time_between_calls < 5).sum())/size >>> y = ((time_between_calls < 4).sum())/size >>> x - y 0.08 # may vary The corresponding distribution can be visualized as follows: >>> import matplotlib.pyplot as plt >>> scale, size = 4, 10000 >>> rng = np.random.default_rng() >>> sample = rng.exponential(scale=scale, size=size) >>> count, bins, _ = plt.hist(sample, 30, density=True) >>> plt.plot(bins, scale**(-1)*np.exp(-scale**-1*bins), linewidth=2, color='r') >>> plt.show() References ---------- .. [1] Peyton Z. Peebles Jr., "Probability, Random Variables and Random Signal Principles", 4th ed, 2001, p. 57. .. [2] Wikipedia, "Poisson process", https://en.wikipedia.org/wiki/Poisson_process .. [3] Wikipedia, "Exponential distribution", https://en.wikipedia.org/wiki/Exponential_distribution beta(a, b, size=None) Draw samples from a Beta distribution. The Beta distribution is a special case of the Dirichlet distribution, and is related to the Gamma distribution. It has the probability distribution function .. math:: f(x; a,b) = \frac{1}{B(\alpha, \beta)} x^{\alpha - 1} (1 - x)^{\beta - 1}, where the normalization, B, is the beta function, .. math:: B(\alpha, \beta) = \int_0^1 t^{\alpha - 1} (1 - t)^{\beta - 1} dt. It is often seen in Bayesian inference and order statistics. Parameters ---------- a : float or array_like of floats Alpha, positive (>0). b : float or array_like of floats Beta, positive (>0). size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. If size is ``None`` (default), a single value is returned if ``a`` and ``b`` are both scalars. Otherwise, ``np.broadcast(a, b).size`` samples are drawn. Returns ------- out : ndarray or scalar Drawn samples from the parameterized beta distribution. Examples -------- The beta distribution has mean a/(a+b). If ``a == b`` and both are > 1, the distribution is symmetric with mean 0.5. >>> rng = np.random.default_rng() >>> a, b, size = 2.0, 2.0, 10000 >>> sample = rng.beta(a=a, b=b, size=size) >>> np.mean(sample) 0.5047328775385895 # may vary Otherwise the distribution is skewed left or right according to whether ``a`` or ``b`` is greater. The distribution is mirror symmetric. See for example: >>> a, b, size = 2, 7, 10000 >>> sample_left = rng.beta(a=a, b=b, size=size) >>> sample_right = rng.beta(a=b, b=a, size=size) >>> m_left, m_right = np.mean(sample_left), np.mean(sample_right) >>> print(m_left, m_right) 0.2238596793678923 0.7774613834041182 # may vary >>> print(m_left - a/(a+b)) 0.001637457145670096 # may vary >>> print(m_right - b/(a+b)) -0.0003163943736596009 # may vary Display the histogram of the two samples: >>> import matplotlib.pyplot as plt >>> plt.hist([sample_left, sample_right], ... 50, density=True, histtype='bar') >>> plt.show() References ---------- .. [1] Wikipedia, "Beta distribution", https://en.wikipedia.org/wiki/Beta_distribution random(size=None, dtype=np.float64, out=None) Return random floats in the half-open interval [0.0, 1.0). Results are from the "continuous uniform" distribution over the stated interval. To sample :math:`Unif[a, b), b > a` use `uniform` or multiply the output of `random` by ``(b - a)`` and add ``a``:: (b - a) * random() + a Parameters ---------- size : int or tuple of ints, optional Output shape. If the given shape is, e.g., ``(m, n, k)``, then ``m * n * k`` samples are drawn. Default is None, in which case a single value is returned. dtype : dtype, optional Desired dtype of the result, only `float64` and `float32` are supported. Byteorder must be native. The default value is np.float64. out : ndarray, optional Alternative output array in which to place the result. If size is not None, it must have the same shape as the provided size and must match the type of the output values. Returns ------- out : float or ndarray of floats Array of random floats of shape `size` (unless ``size=None``, in which case a single float is returned). See Also -------- uniform : Draw samples from the parameterized uniform distribution. Examples -------- >>> rng = np.random.default_rng() >>> rng.random() 0.47108547995356098 # random >>> type(rng.random()) >>> rng.random((5,)) array([ 0.30220482, 0.86820401, 0.1654503 , 0.11659149, 0.54323428]) # random Three-by-two array of random numbers from [-5, 0): >>> 5 * rng.random((3, 2)) - 5 array([[-3.99149989, -0.52338984], # random [-2.99091858, -0.79479508], [-1.23204345, -1.75224494]]) spawn(n_children) Create new independent child generators. See :ref:`seedsequence-spawn` for additional notes on spawning children. .. versionadded:: 1.25.0 Parameters ---------- n_children : int Returns ------- child_generators : list of Generators Raises ------ TypeError When the underlying SeedSequence does not implement spawning. See Also -------- random.BitGenerator.spawn, random.SeedSequence.spawn : Equivalent method on the bit generator and seed sequence. bit_generator : The bit generator instance used by the generator. Examples -------- Starting from a seeded default generator: >>> # High quality entropy created with: f"0x{secrets.randbits(128):x}" >>> entropy = 0x3034c61a9ae04ff8cb62ab8ec2c4b501 >>> rng = np.random.default_rng(entropy) Create two new generators for example for parallel execution: >>> child_rng1, child_rng2 = rng.spawn(2) Drawn numbers from each are independent but derived from the initial seeding entropy: >>> rng.uniform(), child_rng1.uniform(), child_rng2.uniform() (0.19029263503854454, 0.9475673279178444, 0.4702687338396767) It is safe to spawn additional children from the original ``rng`` or the children: >>> more_child_rngs = rng.spawn(20) >>> nested_spawn = child_rng1.spawn(20) `1Ay 1F!q 1F"G1bz$aq 9AU!10}3 )^ :Qd' "G1A  1b#1F#Q 3fCq r"!5baq  1 bq6'r HAQqas!axqs!515S%QB ##Qa :Qe1 t1F!j 'qbQa 4q5 QfCrQa nAQ Qhaq qa "F!:V2Q aq qa%Qat:S:Q&-Q%Qat:S:Q&-Q qQ qc )1Cs! "Kq#QQE%q3aqqZqrAqqAQc#QQc t:QcQ/t1Ba "A uCq)!QE%q1qZqQd!1E!4q2?)1z Bhaq 4s! !5q t1E!iq 'q(!2!;nA ]!85 Ql-qP]]^ Ya IQ y wcs&1j!LA!" wbbj 4q0ZvTPSST xq"+Q 2T'O3bAV3a *AH 2T"E"D *AQ%Qa'>at:VSWWX&.iq&-XQ&0 1E3+ | t1A'qJfDs!uAQQ0`-1h t1A]!4zt7#QuAQQ0 *AZ t1A%Qd*F$gSQQ0l+1r a /q_AJfM'A/q_AJfM'A 4q "!7%q "!7%q 2RwbcrBbQa rQbaj +]!1 +]!1 ASU! ASU! 2SRs!3cAS#Rt1Cq z1jt1A/qJfDsRS55Q28*"!H !1a /q_AJfM'A/q^1JfM'A 4q "!7%q "!7%q uG1,bavRq-QgQ,bar1 M! ))7! !"HAWAEaq;&=Qd#Qa<'>at3aq!!8SoUVVWW[[eejjooppttu-Qa 1aqiqe1e1 51QaqJe5Q,bavRq Qa'}AQ  U!1AU/!4za234qq3+=1@ a /qq/qq0/ ]!7%}A\a{Z[ $AQ %Qa $AQ vRqj vRqj vSj 4q,AT6WAt1t1 Kq 2T"HAWA *AQ 2T"HAWA *AQ 2T"F!7! *AQq!4AT6Q#4q#4q$D37 ,%Q@ t1A]!4zt7!(!9A[1 Zq@ t1A&at:V4wa9AQ[0j[ !V t1A'qJfDq(!9A[1=;j\ t1A&at:V4wa9A[1Jad t1A_AT6WA9A[1s+Zqb t1A%Qd*F$gQ9A[1A@ t1A^1D &G1uAQ[0az t1A%Qd*F$gQuAQ[0V(QL t1A_AT6WAuAQ[0 'q^ t1A&at:V4waa9A[1'"!J t1A(ZvTat1t10LG#1~ t1A-Qd*F$gQ[TE[PQ0 2!P t1A2!4zt7RSa(![1F!X t1A'qJfDqaQ[0l&4AL t1A*!4zt7!9A9A("5QE1!F#1B t1AZqJfDq9A9A[1!1T t1A^1D &G19A9A[1lE)R~Q\ 6! 7#Rq 4q0ZvTPQyTT 7#Rq 6!4AT6Q!! )1Fb1JaF t1A_AT6WAQ9AQ1#at:VSWW^^_ )1GrD1_d*JaN a ~.awoQ/qq ]!6m1Ja #1A $AQ &! t2Yaqm1A 4q)$jd'd!^1Tr!7! !31A 4rQb ! -qt1A%Qd*F$gQ$aN!Q3c#A+8;>Na` Q Bhaq 1F#Q 86!5 =j#+4q0A y2T5s!j qaq y2T5s!j 2WA 1A 2U!2V1Biq z%q2[*AavRuARvQawa /q_$7t1 +]!1 qcj qcj IQe1 rqj rG1Bbj s!64rj EA 4q A 2U!7&! A 1 1 rawas"A"$gQaca'8QbfBad)1CzgV2Q uBajbj r2^1Bb2Q*AQAU!&bU&gRqG2U"Awb*AR{!"G1A3b#]!3e1('%}A"%q#U!1awgRs)1caQ1QQ;be7#[1O1C|5{,d!WA$AQd*A)4Bhd!#Ry'a"F!6r|=Q{$b9AQr1razCr>aEa{"HA""7qZs#SPQ"$b"(!5;bHAURUU__` 't2S!xqS 1 ( (2[*A &b! &hauC{!$+4rBa ( (2[*A1($j19A :T1G1 #U!1 1F#Q 1 4zSc4qc "F!4vQa qaq 1 qQbfBhe19墮A: &WBcBb t9AS E#2XWAV82Rq0¢!L:K1h 51 1 !6! A 7#Rq +Qe6y !4zQUUV Cr +Qe6y !4zQUUV Cr +Qe6y !4zQUUV Cr *AU&iz$jPTTU Cr ,auF& 1D RVVW Cr ,auF& 1D RVVW Cr ,auF& 1D RVVW Cr +Qe6y !4zQUUV Cr *AU&iz$jPTTU q *AQ )1@! 55V3e3fCq rqG3auAQqF2C<( B6G{RSN 6! 7#Rq wc{!1$Fat:U[[__ffg{!1$J!4zY__ccjjk Cr wcz!#GqJV\\``gghz!#K1DPZZ``ddkkl )1Aa1  !T t1A)$jd'9AQQ0aX t1A]!4zt7!uAuA[1l,ban 6! 7#Rq ;aq >at:VSWW^^_ Cr :Qa?qJfTXX__` )1>b1-aAp qAU!3d%t?&0@AoQ"$&:!0A :Qiq iq0AqA0L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@L@LM@L@LL@L@L@L@L@LLL@L@LL@L@LLLL@L@LxL@L@L@L@L@L@L@L@L@L@L@L@L@LpLML@LLLLhL@L@L`L@L@L@LPLXL@LPLPOOPOOOOO0P POOPOOPPPOOOOOOOOOOOOOOOOPPPO P@P0P POOPOOOPPOP|SSS|SSSSSSTTSSTSS|T|T|TSSSSSSSSSSSSSSSS|S|S=US*UTTTSSTSSS|S|TS|STHRHRTHRHRHRHRHRTUHRHRTHRHRTTTHRHRHRHRHRHRHRHRHRHRHRHRHRHRHRHRTTTHRTTTUHRHRTHRHRHRTTHRT;X$Z$Z$Z$Z$Z$Z$Z$Z$ZU$Z$ZU$Z$Z$Z$Z$Z$Z$Z$Z$Z$Z$Z$Z$Z$Z$Z$Z$Z$ZU X$Z$Z$Z$Z$Z$ZX$Z$Z$Z$Z$Z$Z$Z$Z$Z$Z$Z$Z$Z$Z$Z$Z$ZX$ZU~X X9B.? * ?,|l @yD@:5/?@@R2B@96SC@wz*E@r4dF@OOfq]@Ob^@+NT_@ݭC#`@~{`@kbba@YSȐa@n b@1Ib@5ca c@c@ͦ3 d@\>d@nz e@s9Je@FGGʪ f@yyuf@IJC g@Y&g@oFh@·h@aQL i@ai@ F~x*j@&Pj@7k@!+k@VFl@ l@tVm@pZNm@k9ihn@HQOUn@a,~|o@b4nʼnp@+e Ip@cp@)Vp@*q@6Gaq@q@>m#FJq@FK.5r@b)C|r@Wrr@V] s@rRs@GIqs@ >6qs@jB*t@ A=rt@fIw|t@d'-u@X+{ Mu@# u@ZGDu@;#(v@b%rv@iv{Իv@w@Ow@\&әw@}6-#w@h͙.x@k?7yx@–'x@_*y@Yy@1*y@^TTy@,{L?x,>N>>q>>>j>>k>>Π>>F>>>'>\>#>u>J>*>_F>d>+>$>w>>>JK>y>|>iݿ>>I>;>ʾ>t>5<> ~>>>O>>>~3>T>ե>(>g~>ՠ>G/>>>F>J> >:n>bԓ>Q<>>x>~>>>^>Ј>D>l>1>>%>\D>@|>?y>Bv>Hs>Qp>#^m>mj>|g>md>a>^>$[> Y>=3V>[S>P>M>J>~H>UE>B>?>=>S:>7>"4>=22>T/>d,>m+*>m'>c$>N?">,>>m>t>F>>1*> > >Y>>ʗ>>I=_={==^==&=_=g=='0===P6=˙=\= s==d= =yo=/=6=.=fЍ=x=i'=܀=a1y=p=xIh=_==W=TO=G=>=N6=.=&===-H==<א<̀<<<.~2~~r~դ~Ƭ~N~u~C~~~~k~~~~~~t~~~6 < :#%](*.-z/13579;=?EABD:FGNIJ8LMNLPQR T=UdVWXYZ[\]^~__`;abbcod.eefLggh~~7~~/~7~~ ~ ~w~G]~>~Y~,}6}b}|O|06{xI4_h2z3+3@3aQ3i`3{am3Ay3i3*353=3r333|ϡ3ڍ3+333^33׶3iż3-¿3c3%3uY3<3L3gv3;3k3-3$3!333P3P33<3p~3չ3^3J3I34V?4=3@4A4A4qB4C4D4udE4-CF4K"G4H4H41I4J4vK4\fL4HM4+N4aO4O4bP4ٽQ4R4ԊS4crT4ZU4CV4-W4ZX4Y4UY4Z4[4(\4_]4^4_4C`4va4alb40cc47[d4~Te4Of4Jg42Hh4Fi4Fj4Hk4Kl4MPm4Vn4^o48hp4sq4r4s4 t4u4v4Cw4x4 z42{40S|4u}4~44v4@ 4L4>4ق4v444lV44R4F44p4 I44"4_44Ќ4l4L4`4ԏ4坐4y4ݖ4%44r&4k44(4444.4Q4N4t44\۶4H94̻4p44~X4w4p_4~444?yjD?l[T?w'??o?Wp?xI?-3?x^j??N?R:e?4:>?l?*?%z?PՋt?4?e;?$"?zaWF}?Gz‘B?Oq1? OU?ߺH?7a?nV,? K?Xhw?հ<?Vp\?m?)?zP?ZcX?*;Q^?#*'g? U7?e&$ ?jJo?\Ȭ)?L&?FS?leZ&?g ?NIO??xRr!?P_hy?y6IJO?_5%?[X~?1>?bU?+À?PX?5:pɗ0?8d?;U?J?͓?)m?ېZ]G?/|!? ?iT??Wq?PF9 ?ߓ^??ۮY?3???i?Z8o? O5?ٸ?P?R9?igP?La;?L?!ވ?%o?{7=8?Ҁt?DvC?6?=p\?;So&?mj?W?j?$O?z5Ѽ?Ҏ?C|P?yh|?%H?/ZM?f!w;??>ǭ?MAz?G?y?.?P9կ?TT}?g4K?#$O? Y?BM?6C;?B"_U?~t$?œ߉?52?Ҙl'?DɤT?<(i?qE8 ? Uī?OQM?o^?Sq͒?Gط5?zx?1zd}?:R!?Wg?~& ~k?=~-2?ZҿҶ?'|j_]?it?[?8R?uqb?#h?z|J?G~`?\!>?GF?vJ?l󈬚?5hȩmE?㭍?-l ?uG?1i%?調?M?e*|?zè?^V?4<%F?B}u?c-@c?n? R=?Kr?*}T#?,"k>?R) ?K{o?vaӽ?命8? t;I_? h?3xk?3Ӻ?b3?vZ9S?LJisk?M$a.?ftW?+ ?"@|?&#?p>_?1fҲ? DE?} ?/?%,?0?5nl+,&?QG?b. ?,*(>?p_8?cU)?h*?'wާ?dИۦ?ԭ<ڥ?]']ۤ?ݣ?=|?j?.?ĥׁ?u? ̓0?"NR? y? ڥ?d֔?^8 ?0`4I?IrO*?O'?x A?B?/)?7h`|?] ٨v?p?gC_e?T?j%= BT ~Q~U~KD Ga7\%aFOaSuzpD(|Wc %WM$ t`K[oT`gtSwf# Wl`0H7[z1z(zK^2#9MM0MFPrOxS왎2ȩn{TH,ҭ^p .]M[\}r;/4d6dcNQp.t@e$oX%L(xo6xu{fUY>9>{ppCBwS(:5^dܓAN}8) YfHqն&|s f2,2Ztզޗ .n ZR'ӯB)[l@u Pҍ'TȈt(5wI'L/$;nXMØT`OArW,+jtȳRfARnqӊ<KZW$eKs) 4<=>)G'QA@Y.(5bX jz>lq{2Xx{~JH҄Cc`Qz%~ )Q\HsrUb'Bkq-hnק Ψ;3Kd)P^٨Tv$Hx"$ 5..&$ŗ: L< Ŀk<4xV<=A[<'?}y<NG<~;[ ?A?Į?"?ʝ?G??i>l>7>>*J>>>^>>F>>7P>>K{>>>u>;->>>|>eO>4(>8>L>N>ȿ>>>>·>ε>߳>>>0>U>~>|>ަ>Y>IP>w>ҟ>B>e>)>~>_>>C>{>>J>﷍>X(>'>N>Í>x >b>x>>!>^}>;z>Хw>@t>wr>byl>i>g>Sd>3a>^>]\>&Y>z)W>T>P R>O>L>5~J>3H>E>nC>@>VK>>;>9>07>4>2>>0>p->+>s)>7'>%>"> >s>L>*>= >T>T>4>>y >ϣ > >>L>>l>=+==0 =C==8==hp==0== =n==|===;=ں=Z=o,=ް=ߗ=.W===%==r=W= C=4=M,=4*=D.=y8=H=~=x=3E3lM3FT3/[3b3i34p3fw3&~3[3B3ψ3g37!3>3T3d3n3r3Fq3j3_31P3r<3$3k 33ȸ3q3|{3P3#3C333dY3"3+3®3r35333x37333p3-33731b33l44(44h4C44 4` 4M 47 44?4nB44L4 i4a4T044542`44p. 4!4"4i$4%4@'4t(4>*4+4,4j.4/4'P1424):4454&)7484c:4;4$=4+>4@4A4KC4vD4B(F4G4:I4J4rTL4M4GuO4Q4R44T4U4EiW4Y4 Z4G\4]4_4:a4b4d4Bf4\g4ji4bk4m4n4p4\r4}"t4Yu4Hw4[y4X{46.}4 4q4a4]S4F4N<434,4+(4{%4$4o&4,*4'04m84 C4P4_4q474{4w4>ԕ44s4<4d444$4 (4a44lߢ4$4l44x 4_444{4 4EP4±4{:4귴4);4nķ4S444<*4տ44A4.44ע44f4RW4R4*Y4Fk44δ444444g44k4<444y44u4_45555@5ó 5 5]5^555q5v 5!5%5V*5s/5;S55:<5D5NO5^5Nv5?7E?P?'{{?*!?bv?mU?9U1T?/v?x]?&1$-?~ n?cK[!?I?\Omg?f?uLi=?sڂl?x?Qf?ij?%ᨯC?+?Dܻ?z?cE#;?^E#?$O?2m?P"K?>?{s?%;?omo?3;?J9?++?*T[?};1s?HeC?$`?vE!=?ſ-r?MBц?K=?Q}6Ei?7u? !?z}k? ~?@?`x?*?8? Qi?oTC?_(4?ָ?@je?!u v?7Zi?{ ?I?]T?9]??}?8aD:?Yζi?Ɲҷ?r^sSw?ꍰ07?d>[?%۹? Ə{?'HB>?vX#?l1&?:l?磽!O?ލ?&?ڋ?ﾀ+ j?A1?N0Z?0H?}G?(V?5$1!?pB9 ?b"FS?)vEW(?vG}rO?~ /? {^?Z? ?ބS?i"?lR?3Sn?>N?Ґ]b?,|y2?jG>?TLҫ?~>\O??@YH?/֎@?9O"H?>?1 7?8?Ox?]4?5D9g?r|?>ܸ$8? [B/?I䠟?O?y%d?bPޱ?c?PR?j?F}?9(Q1?c?(ڦ^w?0U^Q?1j?T ξ?x.BTv?Imb.??6YJ?)ِ?\C}?%d?w?SN?эv?pa?,Q&?@oű?SuFe?PV?;?I?viׯ?4D?.g?X1Iα?Jy?!dJ?پz?j»?8G;?L|{ʎ?mwn?k9:9??Ry?A&E?U?Ŗ<?k&_?G??~#? V#?_?S?Q| z? Y&?$?htQz? 3Tݜ?pXP?N梚?H*g?gS(u?1c?w@rT?Q=I?QA?]1%? RD?lj?W'n?-BU؊?h?t4? n?boQ?qvi?_)N?]tQW}?6H#z? 67w?"Ηs?C@Wi=q?ḰXl?f?$ka?%> T+Y? O?K 2=?Ɨ$'R~1[} NVeΙVn6nvK zicp%E tQ)2U1WQ9Lin?23F:L"3\LQ V f[_rWDdx h+*k2=Ko:qr Mu\x?A{FS~8;b=ZV`bBtu9=JE>XدGwdO 8cx AFẙi&zqVYםΡag6 X83:뇡koɣj_ۤ| Mg^ݧt|Ψ_ΓXp2X^ttH蟿W;ޭl~$\z[߁İPp:J+N!X ɦ֬ ᆴX7(. Ɉ?5}h.G{tr&oya=cA/˺DH0⤮<)9O@ᣩTrVj֋@?˷dsI^i@(0߾ta&⊂l1EA1T[n&mi#d)B}QJwt}B < EOvpc/F<Ң"Ae އ0~ Rfq(*QtH3D@M`P}hwx%ƿ8*JG+[EliPIw+ E>ҙ02yΩ4A (Nt.Ȱ--̕^&܌z#;ޖu~g6X .pmF 3n bH޵LaEZvpR(-x_b˿ӰdyQӶVg<7܆ut7$MH𯋉ld"rqտH)݄ /0 wپ}2}K D5z&R cM,}uc?Ѡp5.bJ3ʸT[vv+\[U@ضBi"7oLeiFγ>SR(D2Z> B0$y1gWr-ެ @樫(afoeW-|&aY +M]d<A]X`<+M[Ij<[5q&<.8eG< h#ឪV <;LC%K<ꆭhAޓ=?;\ JNx [ (il8$o\$yp(%q%zt(&vP&&/&D'h''''S(0(`((a))8*d*.pl3P4>44( 8 X H( 8T (| H  8 H @ h X X X X $ T (   X!$*T*x237x:0HB`C(GIR S TX(TlxVh^_h`hzDX}(8H$xLt؉( 8P0\Ⱦ,hH$T%APT Y48^\cHgknr$xwL{tHHtȩ(X,&\])[ PhxHHHHpHxX(x<Pt8xx <h(1h33 H:H :h < h= x= =!@P!H!J!HL!N"hP4"Ql"S"T"h"##ȴ,#@#T#(#### $$x$$%8T%t%%Ⱦ%&t&&'(((x,)@)`)t))H)($*X*h**,+\+X+H+h,XD,8p,8,x,h,(,h -84-T-h|--X-8.(0.X\.../D/8l///(/(/ 0P0X0000(181XX1x11(1(2X22(222 3H33h34h'44X(4(48)5*\5,58-5.580,6h0L6x0`6161616(27x2 72@73`7(47x4767h6 88d888:8;8h;9; 9;@9;`9<9=9@9xB :BP:HE:hE:E:E:E;8G8;Id;(J;(K;K;K;V <hXP<(Y|<Y<HZ<hZ<([=\L=]=_=`=a>Xd`>(f>h>j?kL?Hl?n?o?s@zRx $D@ FJ w?9*3$"D0Q0 \h pdp0GEC G G  E (EF  C  I $EC I D  $4EC U C k E \(p4"4P EC BH] A X H m C $(EC x H z F $EC x H z F $ 8EC x H z F 4EC  I XEC  I |xEC  I $T_EO AT G ,EC M I 8|BAC BGF H b F ] 04AC Mc D A ,h EF M A , EC DL{ E ,,$ EC BN$ D (,TEJ P A P H [ ,$`EC BGJ B $T#EJ h I C E (|($EC K G ((EC I C ,@+AC MY F ,2CAC DEF I $44AC G B \h7EO o E ,:EW M K B%4BEC BHi E X H m C C ,CBEC BHL G ,@EEC HD E 0pMUC G FC N,LOAH DH  B <iEC I J \ D n B u K ,(kAF y H q G Q G XlmEC W A |XoEC H H $DqEF  V a $ryAC Q= F $$tzEC K M x H $|tAC OL A $@uzEC K M x H (hL `7L `(L `L `K 8 K `K K `K `K `xK `hK `^K `YK `PK `0K  .K `,K `&K `$K ` K `K ``> T> `H> 8>  +> `&> ` > `> `> `> `> `= `= `1  1 `  `@ 6 `1 `( ` ` ` ` ` ` ` ` ` # ` ! 0   %  ` ` '  2 o `j `e `X H `A `8 `3 `, `  ` P`o `4 y`s`h ` B  ``` ````> @$ (``` `3 ``@P  `@ 0`  `ޙ`ؙ`Й`  p P @ `0 `(`( # ` `x ``@ 1`( ` @`8`/`)`&` `` ` ````Ж " ``p i`d`]` = `` `"  `~ }`}`}`}`}`sU Ps`As`8s `(s `"s``f Gf`Cf`=f`e] e`e `e `e`he`Xe `Re`Ke`] ]`] `w]`q]`h] ``T CT`8T ` T`T`T `S`S`S`S`S`S `S `S`S`S`S`S`S`S `S`S`xS `hS `XS `@S`0S`(S`L L`L`L`{L`pL `eL``L`ZL`@F 0F `+F` F `F`= <`@7 7`@+ +`" ! `e c`]`V`Q`L`E`0 !  `` ```N H A`8 `5`3`.`)`  ` ` ````` ``o N`J C`<`5`/` & ` ````````p `` `P `I`@ `  `Z ` `` ` `\`V`@ 5`1`,`@ @Z@AK  )  o0O3  6 o@hh oogoo6e$p#H%4  0@P`p 0@P`p 0@P`p 0@P`p 0@P`p 0@P`p  0 @ P ` p  0 @ P ` p  0 @ P ` p  0 @ P ` p  0 @ P ` p  0@P`p 0@P`p 0z@zj@zjsPosPohoqPpqPphh*i7iC iM iV`k`rd`k`rqk srp@x @0(%PiI @ 4{B@{2P}Gpi3`l@`D HC I@ 6z$[@p 35 :<P.p@ pHPFXК `xx% |0`0 6( ݽb Opg F =P  [p` b` pS@ `  ^@ "   ~ p  0 { # k 0'_ ľ+U ˾0H Ӿ04; پ`8/ = A @F J N` (k S J 3 ;  @` @ @m yX ؿG`E @ : ÿ`0 пpD) М$ (D PMD  М$ пpD) ÿ`0 @ : ؿG`E yX @m @ @`  ; 3 J S (k N` J @F A = پ`8/ Ӿ04; ˾0H ľ+U 0'_ # k 0 {  p ~   "  ^@ ` pS@ b` [p`  =P F Opg ݽb 6( 0`0 % |x `xU@@z `E x0 DE `H H H D G F 0'DP I PDpJ pJ `J I I @p sD0K K pkЙGA$3a1 *_generator.cpython-314-x86_64-linux-gnu.so-2.3.5-1.fc43.x86_64.debug7zXZִF&!W?9]?Eh=ڊ2Ng:~KLv [(yMD)F,OV N) G_W(Łbj)x >o!tfe hu'sӗ8tlYW6.᱉(Jջ:,fju0/qB"Uep#3f1]F` [Dr[òû A=moޅGSY#JىCŻl? O.3#@: MAV>;W2jAM0בNUDaji:&#[LM 0bT~mP{ȹ3ZL3N鉍5c@[򒰤wX@9@!ul2AVpD۱@ҲSNk-n?K#gXk(V]G s%}z/fT/Ƴ i}IYۘ`< M4S3׫HF#S. [})M`6[zDp b#44CmLd,,^ȧ;SI>v6&s.Mf7ytu&Ros՜.\V֞Xkne9x &RErjnhcS2.nwǻL2liY0jO@% ֐b<-Zj@)׿}' F9Lvh`r@\[(<;T (3L l2[>+~y֦B7`-7R=oHRpL@RJqʌeunzV(܃)SE lbȋ9\ 5Fv \zA+҂au 7gk߱_[9XO@jc 9`c¸mLjr=Dӹը\cǡ̀ӕfx<4&Sz/Ŏ31-(5D-xi* /g*ޕTS cNjD w9?BRtӄ4;5_ՖJfyfW3 % hFq'Sm8\,Z"!c_c8sazXˤh7F؎!pĸ|2YIiwEzU)YԄ$X{cYkaj^m-eVORS|gHKTL_8JCLJ*2B?U+ g86&tɏk[9'9K(V+Ϡ 4qˆZN'cJ +)\vt Qy_̤lӺhr,c($荽=%#t(A+>]淣!N{@sqV3V٦A5orb"qI;` kl4ٰhU 7y@\ݢTͅeixƸ/QL/9#T}*wHĩQZjO-ƕ6C'EG8/+.c6鶈W/k? U#\\ ۯC9IsS˟++dŇ /̆DnˡBdդxx(%|&0 զP-Zf5UtFq 7BF(B0,$\7wH5pށPa;'30Y&Jd6Y!pFhV1dv*t'X,u.m= 3TJy sA# P V<("K.WzMļ U Ej@Eن+2%|͆X%ά'[jnPTPsϩ7Noa]YYUXϷ-[h}w]{q֓-y 3ru+K765 +.(GGN4O xQiG꾛56LK툼I)&vI}e}ّ=L*g;>PbnhȱWdU|UBhe/G0SXKh#x=p/#jE&Y|,Y < q)&ğgYZ.shstrtab.note.gnu.build-id.init.plt.sec.text.fini.gnu.hash.dynsym.dynstr.gnu.version.gnu.version_r.rela.dyn.rela.plt.relr.dyn.rodata.eh_frame_hdr.eh_frame.note.gnu.property.note.package.init_array.fini_array.data.rel.ro.dynamic.got.data.bss.gnu.build.attributes.gnu_debuglink.gnu_debugdata $~@ $000 -{ @3)) 9o00C 33 KOOSo6e6eT`ogg o@h@hhyBooppH/ X X \  6 @     ) 4 $  6 & ( @ 0 ` P   P $"P L1P td @