From 102a972c78b03998f3283a9a82766c26d6539a4d Mon Sep 17 00:00:00 2001
From: Dominique <dominique.orban@gmail.com>
Date: Sat, 1 Jun 2024 11:09:32 -0400
Subject: [PATCH 01/16] bump version to 0.1.1

---
 images/cameraman.png       | Bin 0 -> 38783 bytes
 src/RegularizedProblems.jl |   4 ++
 src/denoising_data.jl      | 105 +++++++++++++++++++++++++++++++++++++
 src/denoising_model.jl     |  34 ++++++++++++
 test/runtests.jl           |  11 ++++
 5 files changed, 154 insertions(+)
 create mode 100644 images/cameraman.png
 create mode 100644 src/denoising_data.jl
 create mode 100644 src/denoising_model.jl

diff --git a/images/cameraman.png b/images/cameraman.png
new file mode 100644
index 0000000000000000000000000000000000000000..7f772d36f02bb0d566384a6c740bac489d321595
GIT binary patch
literal 38783
zcmV(yK<B@SP)<h;3K|Lk000e1NJLTq009610096900000c^UV*00004XF*Lt006O%
z3;baP00002VoOIv0RM-N%)bBt010qNS#tmY3labT3lag+-G2N4000McNliru&<q3(
zIW}FMN2ve+fB;EEK~#9!?EU%IX6bbm2=2Ylx%d6P$cV^!&}hlB23xW`2!qG+1UAO=
zR9#i8fMO^L=qjoiyQ^7DMOQ-=&;|o8uz_N322-~22*$|9#$b7pEZG_?%aT&5Ol2xF
zG9qL6-g}>OcK`6ak>-!+Z{^AuR%XQa-un#a?CJ1tg)2I2f&_YLc8;Z69V{m-TWok8
zL^LZp47FNz!dA4>MgSw41bQvfY|`O^Xga{MfMN_qM}dt_cq{|aMLH{oWdp4dEXXXe
zTO+t9Jb>&N7*QH934j%YKr2e^FpFq&U$7<M78=}Pb~xH#*WwDGquqpZcQ_2H$_=aC
zt$}vfapvaU6kt0PAd?5oPtpK(ge<t>0Y^h5mI6vf70T#7vMD#(1$h|lg6UZ8ec74r
zB`8Np6^a&`B}1h4mVk=xj#vw$nk8f|5hSb*iMi03q!9L^P7Psr34`nmE|E(Cu1o{~
z1*r(eS0zd<7!nBxMATxa#Tr2;4VuV;WfX^CIxIkehMGYz!3qOFEv8IM2$)LCq)Z9T
zGy`D*9zl(zTNs9;vK(!*X{t%15ookDf-F}$!H(wWvH`H8V3cW>X%<=5kcPyXpKRFO
zsMC9nw&14S{{OFl_^L*5pbfhfJ<y`LyKM6U+!kSs&ZVm1!QB{ajKgjiG|lSV(_xni
zo7g>wuosq#;@a637XfN=H7x}qb}|`2Bq;_#RS?lZCBgxN$<7v+5@C{7HtljZaa>!V
zv*qZ53a}dlKqZENifF><LY8A^wPFKC0p<8=0-gH;4XoWsrwvAMZH<+RAaM|up%g<A
zzyfMNuXK>IBYSDU3Q$oC6{5Mg7cf#yC#f+45G8;?IY0!^a?>D3hiR}aqzTF-P{E=~
z;gJD@Y@rAz*cmhlxm^HFFhIK-=4Oz^2F-w@s%?VkW+QHILd4g0xJ=7bi>#nz)~=RY
zfr2^_)gc4KNUA|*r)6b8w8a1t7S$FVLfT;>9A;1y#U#@7vPCcoX&EpOq{=YiphTFk
zM%o4numzi_g32I-Dvh8G*yVx|7AzMA-O++%4^lQ8Z5rtq=!Hh0bp};Nln|ZVDedbn
z)`Hv?9%`Js#~Mz%v5-J#<5F7=$enwVKpGaQ7cTd9X`Bwkv0#t3OnXmQ9c{R3G0d@I
zEKO~yKxGl?LISZbC3p0Qq82v=n<Tme*>bVTZbh47D|fM3>z2KRVZ+<%kpn0ocaO89
zwuQw#SryxQu^e02GDcS+zP4g8a${7w93r`o^IG;mfa$R27DTmy*omQbz!l}gK3|eq
zOQVk>;L;v$6kQmmd!Wk2a-rwu9prAK(d{-|gK*d8?i#E;+}vY0E@Bu1xD~9b0AXXm
zOOtA#)ofM)CW64oupQ}A<KnD<@L2`C(rN}-OAzOcUM<I0nGxu)gkDTmbj3i+WP#w&
zKy{C%WeyDph3a-w)y4pt9EJs$O~Qa%!?Fo<vRkyu@`#2>R743X7zRk%M%pF~(5$5>
zXM-ZR3MF>a3T10k_W(^qse*>e+8hJ|luKTb%?NmRn;=GrLJU<N*;!V`R|n9gJz5Qd
z@L)OskgX0N8j=OEO@wVDHN>R@4V#7voi;?nY#U8<s5>mHS^_vI({#E?sk>+oG~|Y{
zOQSo6j4Fb16b-WwO_LoEqUFkf5eJ(jqoh7-Drb4ZlI4J1BE?!kF)Bu&Wk3+k=s+sI
za;Pgn6jT;V%_3m7gk@->Roy*U=)j1m(k`^j(T9ek%>{<%y8_&4idh*UIV}YsFAM}z
z9pwyXK#X3py<!P#b>>3JE0=|lq>_O}0kv`jk`4!8gG+;mQ~k444uGqvv8W<A2&|w1
z*#b&tR^%p_(TCPj=`Fq<LZ+;cBxhI=9Sme7o#Md=BoXP=OLD{-qB5+E4y+;zmPJ{l
zW1%`?aGgax1@>a+Qew?zyO?OW2+_fo+n}YLB5)`%Mp$msoQ0uC%qp=#dU2B$A_Pz)
zAdM`6+nEMZ5?b2VCh2C2c61AgVo?KJI=(`}e4;K22G|LyV+b~Fwj5@H0-_gUem|x_
zCAd>9NJJPy4hTUy-6k7hDmz$ZNOU6rqUQHy(z5Mzqcwz<8*J0q)W%>Js?cu8YM|h*
zGTk=28ywa0?U*Mj$0j$*U6jl&fC#acHi!`hJ!o%^uNRVRDJW7=%J5>#OL_sUSUV^i
zNfK~Vfp!#x9D^iH1yNE{Jxdijn@y0>vP*UlagjTOv9#x7kK$-F=%@e;*oj6FMVA`d
z0^Mk+P%*l#==o$@1Bq_+SUfA(2H5Rzx62ajAz1Jr)b<uF5G-L&TJg11dvtJ@h4vyK
zFmi-ZTY#Z*6eurcXNl_$`+$%}T{>?Lk#b8w8>Z&-aui8+mx_SK4mgg4OcND5+_s8l
z4Gva!AGNdgYI=`Rh@BO+n{30v4zYSa+f*GRhDdavv)LLM(uzt3fK|#O6;23HS`6WA
zB$C9zQX`i5ss<ShtI%M(+jO_P2SP-&-EhG>92Mw>%-zL-?2FkI1<h715Ee8kRFtz~
zB$WoyN!bp`q|l2bvU5<hIuUj-MbIH55LPWP6;M-{6@D`iv_Mfy1L=r%FA!*Sr=bBB
zSZZRfDK&^I+fadkVRjSYR+)QO58H?yOHf1$%aRJHQ4J1yML=d!!9xWi-K)TmB&mv=
zw>bwGYDAH{1`9d@W%Y^>&}ayb0!T%Rvx8w7qdbozX4-12O}L4i&p0FkgyHCbhR|4O
zt_y&sTPcP`$W4keNE4~{txC{nQ41)!jg*RTt7PxJj^K!HVC$$_cI`@6R2NNSODj;a
zV_)c2H=~6#xGw;*Sa4g&>R1a&ml<L$N69IU)kIU#QDPzbVCKFU?#>SNVGq{2`>cSX
zb}9>6Z!1cjU9lN~WeYDATLH7+5-Qm94{xbiP-#KYBT7bztw?C7V&uqZ#8<I1DsZqC
zMG#!jQZDu?uUr)fF#>^wtGxhFtW<Q0Sh7RvFcPpm|L#Dm*NO_8NH!V>(6Ob9gAI&B
zqf8=g5Hdi)up}A=HpXs=f>CX>Utshu%LTx-4h}-QP_dx_Rud&RR0Gw`?wz(>M>e`l
z)0ML%r>o;@f`mgR!vHOJcEQd@4x)`N*ecCe>4adTD_e9)FlIq(u>mv#cI;UavMg!{
zwC452CW@6uz>RtNl^O+>htai-o$k&7%R9Plpc`D}28~jw$TmSWS=p8hV}XNkSF{Ku
zNEgUr!fI9;1OoxqT#>L=1}GbXh-hX5Tm_mcYRSbW0Fz{~Mj2QJ>j-mmv%4tEd$-zl
z*+!AgAvN4$H6%c-aLE{|Y6=AxB<vX#BGnMj0cg+8DxuL0nr^|0>QcopyLvRrJ>)ib
zw);4`sw7^ty<>E_1Po+baM-3{*bxh~$8Rw>(8YEI2ZfTM4rrhPCTMqc7E0|2#Tn>g
z87O*(AoMzh8!=P?jvR|=aIk6jvY~7^nshbYpovZMLMI3TD1~j;0uB&#FC?Nnm1AwS
z%1q}-GX<<dGJCOo*bYdnc6O+c3xr4@p-?ZR7B0bzElFwNLU2)hC1<;mpoWMhHlo?y
zO#>Y0JxU}!uelLs88F=r!Vwp$>UK+n1}zflD6id8iC7T|g0e%56r+X603)<8p<qig
zL$Sz^kQPRW@FG|y5y7n3vod#(ZdBSZAvIL9m+X=5sOIJvq_->szy)fl1zg%yTStTO
zHL9?g1kB)IvjNgj&YPz?7J*$k1T>?JlsX!U!TI`yfG5;TrW?fExMm<mx6l+}M1!^k
zItkj*LCQd<9BA1v106*|Y*>SGv^hvOX$S@&L<t&A(J?D&qg$cDRv^%hwpz3cGIrXH
zeFUXyRuW0bW-aoVeQtt|BG}_VtW{PuV^?^&2(e9CV|96{X!N0Bch&;b+9g^>u*75+
zb%x$2S%T=SiJ_s`*#U=WCjfLr1`H1s&MM5<)6otWOO-9>&@Mn7JJBs;q_OM;wY;?U
z3b<uwR}!vvv})Z0+J)G7*wY)FW<`9nZJF#tNh#=TTheSsh+L52RZ&X-F;P*57-X+|
z0E%L`HE}(I%Weis#{=x#y4?#lc1Y`v@T^OsX+fMt+fb=#2Z^0UgPn$3hRI<e8ld+s
zR5zN#ZkonOs2ig@2AdX|>bB>7Q-DKz4DP5NvKe5(hT1^KS2va5Vj@ux8ZjiRv$0e(
z44AMNNdiD4y%4=vwn?yK?i@HKAg-~nngbx^BB<g{w3-v=WHig@a+3pY4soDO7j`!2
zHmYJ$Yk;buAV*q5ooY2{I=osUtwtLo8Vox+q+S6woCcja)Q4^k#Rx=GqC$f0J)8p_
z0c3%WR!h6E6ajz^1Z!Y4Nb$HA8XYpVOBn4mBoiqC0W^$MP!s^P1+uFr&xEo8MiPaT
zA=!2V2&{lyy)sq?2~?|>SV3mSiZcmF9)Jj;o8~ToOfeYIw2TffY;ZvyMbb`&5xp2|
z2b36&uhCNIGaHf@sPniex+NFO*wYr;pf=OvZ0s>06umSnsz`X0R8fT(*!M=9TKTyW
zCkr*IU{l&q>b7H`_uYVtMA<C>K+DS@>#T}}f)PomX*recjusIF4FzP-i?JV~G>eTe
zxVAYQWM_mpmYqQx^u9J&4sWaCE4!gCSuRs$a;|#NRv=}%R)8as7N|a0Be4`9AR~KO
zkt4Yz#9WZk%?vqAIYMIgk{J;Sd1X*ibgBdCga|N2A%OW8tKAY7XbDyrG#5v-<!~U-
z1(^tQDWFxCx+wy*S!j|jBKByO`)ms|OXvv)2uAQ2Un!85qC-4=Y=LAz36csBG3ySp
zA*WiTWiX1ApazoMqA=kjDY;X&=l}`B&Zc8DEOx`Jwt!t04BKD>m<0wKYV6cN?-H6P
z4Bd_r?1rUITZNrrRSu&7^<<Bt3;WWZ9fEUo0mJP1=V&*Y)$ui?;bn7_*a9>=QYai_
zK{Pqm3Ro5bWB?uV0jcVR0A!_MhA9QW0E%RaA}1#tiCPT@QVmDbBG8>B_HakTAR0B-
zDQN&;cXqo0VQ{J7sF7`p5?d4J9RRzd(P}vxO*JBh)W`-LN32yr0+Jk<q|%ZM?U5r>
zqQEhhtYtexk%^VTlrau<evD+bb1qP`O&^+a1!J$->2nfMprb8Tf<UXaj#3s<BWqy^
zBq|5I6pIwGF{|BK8UoM=3aT*LM7joo9ks9U=p5~+B6DoZEbg<93&h&e+QG8-Hlo&H
z0KIO*wB5@PVi2^7h;P7&{XRja;%GN_n?kc8n<hr9Ohrj+1sj8Q!5w5*vQ7(t9ywZ@
zlO=SUh3qgyM2R%35gRIoS#}W^1RRZKd)aOE=qS70s#XuqL%Q2Q+foDsAZ!RIgwUYc
zK!h5U#6=R440zcEuK<xs6AQ%l0&2i?HQD5`yHE;838-zLIBG8lMwHcNcI}wu&D`(Z
zB)5jx7Htq+y+l`F05CO9sBxTSMx>TxuFE`P5uE^C*-&~#dGfuWKxntO#=IXjf#GUm
zR0G)E=!6&mWj4a0E(qKOaDix4G=lIbLg;-7c1Tin$JYuNPQfaG0kPmPP^pG`MHCFq
zJy|Nt#EES+(t@$elMU>7{nHL79B4A;;`Op>yU1L$0Eu*p=~*NQCWPjQb~=nGjRdT4
zq?S%ZfkIOg4l`6MIhM{!NsZ7D80r8;H@$2D333ovi4KxXGy-lZfE?IDxh=iyStxQy
z-1fy8*wy3+Mr_Fuog-pf8}0~O2&$xRB;ZOgBy4Bp;Nn`K1lQ_X(yPXk+k5`mb?5o(
z-u$xrUw-M_;qEY+jA#jA=2%;hx?6JYOVdN%3Dg7&o2?;C>;`-6#n^$GWMqgnAOI~y
zmxUom!;ZQDqN(mi<pq~L+zx&R3rhz}d2!}_bruvOTF}G;qSRv6tUXmkC)u$l0FdfR
zb6dG1O~;_4WwDkbXdih0jrj1t{plx;!=M^OJowt*@HzL5LU*g9&;;910jk`itVXy<
z!yR{)Rty3tM4=~*qH2;RTBWMRfIF-mQEa<W410(|X=2`>z&+SR)zZzs+qP^un*?&G
z1XEFga4&nB#OWKSXjO1@5**p4NOx*D=W|xLMYu!luxQ{APyOAW*>JeC8JLANfGdCE
zEn_rXF~B})*g!XG9bxy<mMeAw`?|x24Y$uj)iJV0YRtNU6aKE~Qn*_sY=IWba&u1%
zs5P<}I~^lRjb&F06<_U?m5U;BAUdQF?Z{MTb~+3vDuT&k81NzsIBbTNk)%4=u}iaW
zX{sW(sD;gN`2BzSm$u}^06~y~Ky7b*;S&+r96{7lbQS=HqU>zJG}_^ExMM%C&#i(s
z+2H~)6e<I*$$Is0Rj`&G(Qx&N;&))=VD2&vs&m;pckN4jEgkdjhNtXm{w#ob*{c9J
zuokxfdc*`B+)dEUy*Rj8d&8HGa<{rw19mm&ef!C8eXe8dm@9%d48r~H@B1VKnuWQG
z0jX@yHy7<e#z`S!#jwE+ZD$;SA=gS+A{;Tikim4C27?`y$$f!SfYj>QsgmZT$Uw$d
zgYJ3EV?%eb4Z`d;+IBMt(=}Oju}<Kk2zp^d!emRu*ia-HB9@LK)p6_Re_>bD>KO9c
zOic9I96xj4OP)M_@ZJN+Km|Jy0vC%g)!{(`U7f+6mjz->z@;aC#HQdB?Wu^u)V#(n
zz=A-?a&Sr+tZE<tmB{$U)>3q+0c-Y)4%2<;Fe)8`trpO*(E?R=vpP#2%M>;ms=#JP
zN14Drze_RK=YPs<?{2^ib`TaNV}IY_br<h={Phlj3lJayHU%*q+!rJe3kgz-!we8a
zq=g01YC|1{X%NfcYLYe?oY;|<L2lYQ2sBICwp0joe8tiZuT8NQkHsF5R4p7*jFmZv
zLn|9dN)QquED>a{-77K>Ys(>vi>O$FOr;}`in#S77;W1y88)Ihxqj^LeC&Mt_n&dd
zMllj?N=P@D0_z~KOiL=wN}H{AqQf~CuGCQMqFU^5HM<ZM_R?nQr0b;V?idk86+ISP
z@eNX*P-$k%p#x&-9B79OFIGn*1zbM};f^9S76S-FySbCVZh~aPRRoJdhT`U*d12s0
zZ8B7FOW6Po;^&`!>P9?#sfmakY1BSt9iuU_>;~y#2Uy*MP46mT7>!o9vqlv>hIW_R
zM8{4ydXs5Sbr#Jw13;loVcBiR*JTdI!bG2q&?a`0*_dsxXhtv%5`zbVvjn1owU_}2
zX;Hxr)Rgb4A#!3J`r{vJ%x;4)LNt^+WGrpq>F03co>zcy+1-J_pv?*k4HvQ{1E}N*
zMpT+!8fpiDbSp@Rz#_qEAe5!M$CesMAkZ`+S_mzpq)^lzpk*xsjZ3X;!=V(ppj$#4
z&al^jVzxHsUP6I;C)lAxHMC^3`z$MLO9360xBZId+Xq3iDnlBKRdz_g@#fQSA1t-e
z>}@ZHiHh!W+0xD`qPur*t{O4yxG;8i??CTB4Ypa{s$+2puG=-*;b?b<yN$sE10EC&
zZbWdzS80T;vPmhB15IlS?=e{2rs=`lhyj$*IS611G@7o4QNt{^%kY%`0o#xp`j_u6
zoC+Y2ga?QMcDZ^|-8{VX5$+7fSUNbB5Y;4UtCr*00HY;VCT0V8Ff$YzA%%HXi|EAS
znk-~s2?SsP><AZGq=y0|*#aqNAUzLf8!AnPbLY<7)Qym;ywn90Rk5I&fSOCQ7&{?C
zQ8p!Z!YH<YW;U{^pY6M!|0IhM-He@O%v%|^Kk@^wxU$rcv!mJ32}oUGMiJR*LTpCE
zCe%6M$Z_yIZFU(*07Y}cL+(~h^xGbe2HKg%>`u{awFCk#0I@`bfnc?bFx|ks9gRb7
zyMt{;m;*8_RVpW|<Ul~^p1mnEf<T2{__%-V{llg}5WCyJ{6DL&{p|C`BE%$rnQEd$
zE^2asjgu~jtPs0<{&6zxCeY2@uuK7RM`u%3!-|eZ&{(Y6YR3+5HKmDJ+7CUik_Pf9
z8(57}bT25VCB|+oayFukb&sX#jKSEVEivy}XjD>*F3JaVFYnn2(-YtVqroPcr)L7A
z1e}*!byi-7w#k9Zl|2Yypx`3A0Yinz9ad!*r3+|^sncMW?UG^DIE);JBZuJTF*zHH
z?3VW47kNh|aYp-OJRmY!P$f$3K-Xebg^3YX4+GxT^ic%%QmG-M5^y1Ek}fSNHdz_x
z?>0=FJ})rq=eteO)P(_l<=MrhJ%@433T{(;%29%Q%A?s}3}PUwMfC37<%ZD!&;T23
z%j7zWP<x-1Jv?t`?X-HNoqN^*ai#?*1k{nP$;>n}TLs$)A(r5to!uA=_N=3-;RjcC
zqpJ&h+t8Tghl&Fw*gKAwW(DB+4BXXaXgV6Sp3UEl&pn)Dg@qVq2gL%*B3Y34sZcLk
zk(OPva#=_cXtw9OYY2o}6xk{6NYM-g6#`d#6%}cVuF3(<9(lx4VnUmy>U%rb(T<WC
zV=_C2m#G#?)p!=53aIF4U@AEr&8R2{O_|qc^UHY>5;Q05OfQ#Sed(t+Hu&=6_utZ`
zSHA86sB_kAY%hD%(kh1ylnUV<(Qc?4AWEn(W!c*jQ@%T;8FkobI(ZCIql4{b!flOa
zU<f0bSvz1yb{yGR%N~G53nO}4o3V#cGQB%`>DJ9FQnh1%0qD4RTb2!gMynv?(a=!)
z;NSg(eOl`DnKzt{L@xI{a^+>8dgWK1_KJA+17G;|=O26XmH+UWOK4xaqY>2$qY*4b
zM*-O&7VW{E$c{azI#{D4JxE8@qJ4=DkEwAT%f7glsm$PU8KLlW^96Tl=ih@2#&M;f
zZLvneDETh#l+_H=gxp}5og;If(#S@(Fr5)NK`WLZ7KzXge%p(80QlxN-TblZb2IS#
z(Jy(^d0ZMl{!d==(hofK(igOxT^j3e8J}{$F$7Lxbznl??d2r;t6Fz;3|b`^#c&a|
zYzfoGpooG!70e?Hg8Nu{Ds>FNQbAS0EzM1uFbbXYPP1WWtu-WIi?uY+2s<j9jHY+?
z(j(e(5m5##nT{UHyRZ+}_iD@v7{He<zx2UZzRNV5Fa5gDcp%mnZu7I>`QWopuFpnB
zU2uH&%Pv0m^0xLKmf;FkhfQW_1cxJ7^O>-)Pfr-5fT2OdO;maDzHqeUMz_I=9Lq4H
zx1~ik+*{-j<7=HMLINtAh@_oZ%g{<95J(F+4;b^KN$k~Df=~pDltkDaQ~A6aX{Gs#
zAE?uDk9&Uao9}zuvjYb4>Ob+R=ZxD=|NKQ>z4h<7hL%fN^8L>}|KuCFDKN`G4nYw;
z&rGx^(W33#l`x~SFf_6CMzYy(gmIL*s!I*Qqg74Mp2ip^Q7EO*(T*Ym09$<Xuihfd
zvF^Kyl4BqlBYJJP)htF<bcbQ@5T{U?m9B~+bdeKP$vP<t2)x$!KlFM&`Rr!A<c~aX
zF`oV9kG$}!&ws`jyyG>m<+Ap|xia6>_YAbW-&sz&t?00atc@I^%gq?ALLmzq6hM`V
zQ(nDv4M#6*wg@kSC`MDUrPWSQr2?`>f-$lNNS|$l+dnJCP>wB)5Ud^H0a_b_I9@7M
z8~f4?FL23dKt=C_i$-HtbTFQJ>pmehP&c>df8gP_y!xN~DEOxP<IYoWf9mOHp5H(4
z68-jf^#fo0lTS6PMg77r<bC%(6^A!JeA93_IHH7<z1UGuG~viGG5@i+MpA~@!XDJA
z>ZhEhyJAT8SPV8wwg#|+7*aNkVaA@gz1le#@>oXYhAd7slZV&griT{{BHhfU5j!cz
zzBDVEAUjF`$##qX`lh1iHGt=j*YA1ub8dXbe?0ajVtL6eedzZ6zy34%Z*KjtAO7(5
z_wH1IHv5ZT@}2hj_go}fm8~wSiRrY#;FM+=O=dTjR<lK7gxOd+plFb4;;cNRmK)hb
zk%)2VfoLE>2&2JHZ_5DTi!7rCs~P*Uc`1~3wi%<@**4j%WE;mb?#x}#$aYPgWv~KK
zTHoo;(=D*>TU>wqm*U#qhx0AhOP}%JohN?pOK<Ax&#sSM{MeZV1Z<z(zkZw?7dO2C
zJaPdiBBKg9S|V#QLxM(t?I>{BQyo!LZR=g)pyb>%Z8Mca>}AgeG>8UACJbLJFw)FU
zwH;HY7E^m3h2Vl(;2as-LMS$-gAVc%7z;w|W!LJF7(r}j`^t6sq4%ANej~L%|LpJk
zqfcD^FwP#i@A>r*>uc8AuQ*%R|L(7S<!gD(`|=4dZR*C2*FUE3ew9D`&R_bK-}KYR
z|KU%+?#+8zQMnU7W`(z<;WEpK(Ew#v&em@Xjz#-27+W^st;OmbMX)@8tp=Ta(&&r?
zG%mGV4UMcO5Duc`Lg*yYf&i6sV6oXo?ov%!CIZpw@EOZV9L!K2oX3Cg&vv8L04U;n
z|KDf7`*}S4={H~U@>{pZGFp$_;_A<N|95u;Uj6W=TpoS@o#oCCzwsYD<NwR~Yaafr
zN8-Le;K!pHrQ%S_YN((cFkA(~j-o0C1dps`Gz)HZ4DLb)S+l6993hx6mU4xSL^O;+
zsnsetS~ZL7Wx!%Do)RA|RY4g|$f#Nh<7|t{P*BId17)f&Z1uCxy)%FOSIS^V16mS%
z`a8ew6Mu5t`(<ZZa^t>TUFx~xdiC(x-?s;U)5o5F)%!0lUU|O1;-&xo<Bco(V=rAE
z8RPyt+A>;T6Al#Rk+bvB$hNJ49Wl%iy_6uJI(i<x6GG^vPwEKPY)s3L3z5+&VKG`o
za4{00&2kh1-JOXRa^A%pIRdIJkZs2n;j<fWe}}Js;8>6@pC7Mv%fkkuNRWtoUh({m
zrTSm5FWL7?WG;i+?me&PAAil^_L^M!zGvU@$yfi<bD#U0f7j#R@C`qUQmpUz>)-nS
zx${^4$CrP7(3KuYX_v7ld*|T@^>jbL%yOWhJ+fP4I+mlu9o?+6t%G44LaizG<p^vE
zCIzFRDl$e4TFv%sE87UNMW9Cztcd2W20E%qN9XdcZ~F+yXjs0O4}KCo!8WAJkG(V=
zd;bfH`o4!h^?@@ob{>}A^uqgYJbcE(rSn_;$aCwZClA|0@BZ??{@%a;36>Z9o9k}_
z-&xGzURqhPK?s6i`XV&4iKUGWlWH6~-Es>{L0dRR7!iGpDL1c@+_7|Ekcd34@eQyv
zrI1mIQVao%5C|0mVuQ`_5W<BlFQB26X)V9<XP@pi5lx^B(^FfI3Hp4RuTobY`k2{g
znao4(%c>=O<MBNF!|(lppLx$ms{66$KmMK<w%>l=-?{!&|H=#P&%g7lmh$F!(}T2#
zCc6VWa$e0TYYI=Ssjcd?i2{3m$B;Wr&_jZN=v@xLjbXy(&Lz8<y-P3#duLb2Fw(uC
zC2mm#FR0lJIZ(*iP&CvB-}&EOpuk4Zn%<Qt8tO?a<MLa63O?NXmY+Kw?tgk&FKNGL
zz2$rNThHr}{DvQT;}3wBj~D;hwdZy7h1bQyzv8d?!XNq@-{4EjV?W&&-}tFlsvJ=*
z_xbROVZj~cUT7XJJqv|RR~sym7Hl=TTO-SG)Bt1ban{~zDH<JcpLJ}>K#xPq>$bp_
zEO42mUdkkVMl@*WSSzR|?kc^4b@^{Ua{~QK(=%)jP5%fq06zZzdf#>)mp}FThhP4&
z{b^P2t@7hnZyfRD;p1QUWk3GFQ@eiOKl`0e<uCG|zWF~2ANiktW4%0m<#&JZ+Lf2Z
zk%td07ULqvJoGC@1IR}E&K<g%MF_-*b~%P<V{t1shYHc?7FsMUO)NXyTm}cy*+^1d
z(d_B^k<%SbPlkWQt}Lr(zu6$YPeEQT@Bg8KC(k8-Gqhwfz(muC=dS&s|9w59KJ&5X
z58R)&ANF>(pH=<%^-q2I+3R+l<(*G^^!(2~=<QX%yaezs{_w3&e4uaNxcPsK-+0gU
z2hV@`)@$glrBE8c^is5%U~8w`it3#9`OdM248tXOHc$q=923Wg8e}y)<zsX>)s1Wx
zh}nTnspKwn2tbUq7sHy8h?#5QD9KW8-+Cud_H@5D)iemQPcjH>hmU>wm;PM(=AG?G
z_3^3Waom33T>YL$`rlrB@>>cDzW)b5g6HJt9{EX<{FO`i;H#G9RUiM@pE%3q_SGx5
zHFRkUL$=(+wA>bQ3%l7MI?M*6rm@try2a)m1)8G=vgzqs1WOCq%9s}(my;@+Tcq1^
zC)_l|7>Q0{P_wvhQxXEwcRjI`Rr9G#jU^^eYZ@fr<Hs+)@T0eH?-%=a1Tdbfe87kO
zg~wj|d5{0%@3f<_Y(F0CzPR%<XKWs>eBJkd&mZOWfA~A^d*k0fe$`Lh{F#3byy{O!
zGg{1}b7PLM*B<UfOb6IR!eO;qBL|r*WUcPXT57?5zA=c22TUQyo`c~*X#geQ)Ne|S
zu^?uOKs40d8j{9RhhtiR75Nk2y&Fc>3GlTz?Ix268z4NM!})LdUX7bKHj!{WzoZ5p
zc<AY$`tWC2&LZs*YTb`9&cF0c_x$9=BWL-|pYq^+m%izW{)_tbi+uEz=XVIWqR3iH
z0%BKy+_KUth~?ZFQ0KN(G+5E4;cPF%su%_*k_dzmg+nR~0Ub@j46zrQEQum)!d?p&
zxpWZXo)9`#$AA3x3xGPQTWO+CTgfnR^_l_Kzx&dAU-rgdJ^+h8M#JM{cV7PJOYdJF
zfBSFzGBZ->fmGFl$4etV^PRu?iJ$(eFUn6mJaWx(bW~m3j99!AQ$QdN7o(d~y~u10
z!Cq^mM>rs$fq}Bk*g4AB@|4k3Zn8=_3@47^Wy?W0xi55cbR)!e>y#NY{i|-YI8qL8
zE9>whf9uW(o|=o^CQ7tn3&j8%@Xvkh%JS?huU)$GvCB_APaAdpqR;PLKIO45DOVge
z4h$d2M{e8#FP?hcyY3wS=^wu7KY01-ANvdZ>Yso1qwg}#zJfANSsP$!kBaS3)0hT2
z#$v$MAPgeV5E`8oP1*spWY7YOYjdQLBc#%hh}NPJAuG6ZDsz%UokI+nV$n;9@*%wb
zyKlen3G(L*%`m~Ksbnkg?8E}MpS$(1Klc?c|ImuF2VR(J(&Mi*Z>?v|CcKJSsIiYk
z;Dc}c&?mp+qx#Tyee%a{T>JhH`8%(T>mbl+?J;t)kuY~{+-zH8DV8y8pb}-PMp$T~
zu`!HkwPJPm<VeiIf<5CKTT!A{JAozwF4bX3W)lK}!t_Sh;sBenm*q$P#*rp4|Iu{*
z%{u^q0H7v1`k`x&UV6~iwzccS)olRd+LIr9;!_^!qW=Ba&D)WJ_pUn&FSbV?{}}tV
zHy*r#dw%t0SDsxk4riAh{UX%)D9>-HE1HZ>(UjBxB}hXW3t-J2)S{@Ntn5-RfD)F8
z9KuLBDQ^OzCTB;QLDbQ$yTbVr$I8u7VGgw@to)w;?;SKGH*li53hbF81DG!|V7&cx
z@A((J{DHH5xjIF@1pCXczVcHx)4ktRL^sHbtDpArmp!}v=$&QjFTMOr|JU0+*1kBa
zjcq~u4MEK=%y1ba?PbirhbrVQqgb}LKs3aiw6i*<&Yh;QG_8(WmeI_fc8wUTVM|5t
zSuxdDQ9$m?4(_SgR~7|%>9zjh_m9c*sBvQXku@=YbN;($xK;S0U;46-zVEp@>WPaf
zhl8<x%{E;F{rD@MMw0tOulVSL-~IRgY<>EV=hEN%UE4P!&{&xgQ0?Hhj6m#~J)l|a
zEoAO;bo2<MMOaX47wL6(4g`uqX_zufwL13H4#W)Ma&O#^qpR(zD8q%}u~qGRZ6Gja
zF4EMs>+wJJFE^9Z83l53pTo9|2Hf3nC=>S5|NLEy%a=a-(eul75((HpoN1`(N1uEp
zJ=o(l<Kz15r$6B}kHb<wd!rXwWTS^&A|M7~MbdDnf)U;76y43%v|zi3B9`dxVgN+T
z)uCKi&hWX#tPTVcF}Xx?Ms&5jXjqfcjgFDT>1gYalU}wT_?OSmOu7>Rq9#v$ut~Q-
zPX54wKEL?LPyf2*1J`bU$@LG!ewP;h*wOSz<FkMG>HDuC9U=UyD}VOv&P_UbgSd6?
zSgOJm!^MeJHFAXaDTIiMHaJ?42`{@U#KG23RcKJ&14UH<MYoP?1$XSB7?hEUMjoMN
zCy1tO${pP!C^|{YOI6PDA@b&5dq>aD`{JQD)6W3R7=X!0gD}(wczkR;_3m50`CmU(
z<Fo?^y!|r^s?1M*;NSkicaN^yx9@qwqVTS)zWg6wvHtt7z?d9!RD;<bX_m?kLv9Yv
zWCEMA(S@*wT6?<WE!kiJJ@q9;dYvXUjTu%jl3?!Z5sn2$9B05!t7GOB2`0f^R3Ry?
zPyWRZj=N@h%nGPa0KgQ6@ze!OHG>7!QxCj-f8e|JsmPuCh5zg==;lS|fAwwO@-V9(
zxc8U0EJp26Ki`k`k6r%UqfxquhKd{LE(c0Ip98QPK0$V1L3?1(?B1JYj^r?CwoVY5
zhtOn=!@!VR@wGPTaGq>t#2^JS%;NL_MaM*TDIMPPmp@ojL_06O?lxx&aDqYRy$FQ?
zx687>=}Uh8y<0@CWtspoZhiI@fbM?P&0pC^-MV%EuiUt}{=%ny=B0PvxK-Y+-Foti
zAy^2=n9E#ZAkmSXR(er(OcNWYWu3teQgk95ibh%%2?il&u$Wc{9f+?L!^Oa<!9>JK
z&a_YW3jv2=h|r=R`)ePY29GK2I4%5>24I#&5CG@>ufN{K0-P1U=R1CM%V6VDPoO;Z
z{ont(D=xAQpLylRy)SuS@#Qyu^Ubqwsvo<7M3?L4=U;L+Xa`gfjg$X8(e9|89%K@*
ztA$vGno(hh0+#lm=268?3k<hU0c4gddD8Ay0xWTMiY;~6S+#hI&WM)S-}9ZjK`=`O
zS|qz+-F*hrW=t3(_XBE%2%^6AXC2W<Zs&1f5KTP)ZGZBKANpgj{g)44dh;7PqoU3}
z>5KBp{>wKHe3k|dR{$~G@F``LaFj`?5>A2E)Yv4wn`%|E#TXFP1uorI$#59^6l#+_
z*1aHQ8jWg+<!!OMmZAyo45OxTha<*Xp{RfKgLT^ZoQa~-E#MaJDj#gNVAWGkEf3wa
z0T=FE`<6fUVQF(}ySxwZ@+ZFf<(GAQ`lI))|LSA+e#cADPe00s-uk=p_N(IlFZ^&_
z-a{Yx;|G+~M)g7=+N=OMgMB=cIx9vN(0~vPF5cJ8gj*wO-OW%DMr=z6CCe;ohu9vO
zKorf6TJI9s9Ei$6)zUCLk{Fcg+kc`=q9%(#+p0K8fF=QG0-iu~^>{LO&=_&)|MAX`
zidd>v4Djk#ef_;p-N+TuI9~ho<^O*B`F+oE_Ui58BJO$RcijdDue{+C23#r!msYEk
zLrp?uPHE%Lkg)@S!gQOC;F0tUvdANCw_VhcgS3bCu%lQpjLZrT3Jqw=XNv4>&yC{~
zGTaf!+yDE|H0Ckhf?##K5vFnXc<0@%AlFO)F<@P~@huj?R3Gp3>;L5beee9>kt<(z
z<B4B>@_$LxTmHv;MHj?=?oEFQp+4O6NbW_#VK;V&L&6O1nqf?axlgA-&A{1RrY-KB
zFj$@CKwFi2*z$xdjNM@8j*+YPWIJcVjz$)kZed?-!R)2%q;mY#p96qA?Rjbz;Pxqm
z4u*^gM$wq5JlPF`<DcIU;c$)iBj57Y@A-)L=BMeu{r;!_+%q4o951~5zqp44%<XUf
zdX@t>-~Z?!xP#p=7&~N4qZzUZ7Re?D<!ZC6X*vSHbyiMG8ASv%mZC2b!A2a(*nt`O
zHUlo54VDH5$r3>}HG<9|63mOg`Hq?81<jgo9+=Hp2;I%~;HeCl`n>M>!$WTc0FGVe
z!{7QNfA0muUf!Tzee}V@r*Hkzw|>EIJ(M#$8vnQ3pL9uq4?XuyZfO~GG-d*Cj9`{?
z?DD`#d^cZ4bKTGqfnqOcFWXG&*r)slXss43l<1);)y+0rd(?4hu?ZX9odYne)+x+l
zsc-+Op04`SVsFh#$9J`k#6*oxTcH7H<Zb}<K~Z_!t5XmD<#+t!XM?nlf17{&(RY3B
zmy3S;f3@J3U%4z{{*nLuO&9G)_0UTm9Pln#rDhvs$x$tno{>*LH<v<t8K#;oThR@;
zbC#z$HGF#*>JAwNPkTf|IV=vk(=J8H2Iw(^cg<c|Oy~Z$ern8?hnRkM+5}F$;EY&0
z>47IA0vO>FK*V6(ss{VYUwrT1dp0=r4Y=<$^#%?c|L`XZ|G}S>*7$c{@XBY_!_J$p
zh&E!_MLK5yQVW8LHD5!Amz_mu3|25W-9H}7JSK{F@8y`O3hLami0LUJm=2X((-A=m
zXmuKB2ZGbSb=d#r+e$c@dI3;<iaM5Y^6X=(mrl<>?X#$THU_JDpIPYV|Lc=~=V|F~
z>u~LGN7nw)-zLU4@m%J|?tJpwZr=j6)G*xZaovLnU@cIU<qqeByH<DvqTTAwlE&FU
zyB*zRcdr}Vbttq3h)M8lVW4pohmiEXbf{y(cr&N203uDe?cnO~c-u}-J2M)y(YI%W
zO7oO{P`s;vCJfh2KskRo_6B3%%Ad#I{zxlZ^Ns%P&qb-9_-258wOQK!;%naU7YiUc
zSlBhR13Lh2>e#wROld>A*JE_H#<X)4Y!Ab8l7S6iRJNm3?h$iy76X+H3j?jxT~u$X
z158sU*3GVLG&1DB`%?qpbip)d#KR0*s=5A~L^P%fZ5{z`n7xr;R5zj)u7Bpzx4hqU
z&u*J|=A#dfevW7FIeYcZTW6O8|KzQI`~7btfZl!+%pf|phCsVG*Yd<%+uDO@42Max
z`_P<cuF5P`s3%ebR79cE!7=e-qFX3ySUUi^yiGS94Qfz~eTt<I-~ZO#pxC3f<96Kk
z(XhL5cO`quwW;Z^;fWLB-HxydO755cpO5^_Uo_}>liKd@hWs?@;{W*mXT0y*Klq>j
z{3rd%TL7Uie#?$34t2S`_l#((Q7|M|Pa?mYVUG=B8s0HOWeS=}mkKRLp%lx4gxjQ|
zT68$Ag(bo$IDnn<v}_sO?4nXX|6KzL*)!9=yQ{kIZh=`3%)C{@6nknTLo;QVT`e^u
z|LafsTR(#(lkMhe0RCC&zOT!F^QPbX?uRd35+C@hpYvzmX2kj7lYU1tmUX8c3NI#&
z(3;smw_~;+p#e~l-J-2%PzKgsA~g*kLzT;-8XZL|quHx4wO8q}j;1V5R=qgpZnt1+
zTjJI~xZ@WS>k~6@be{7k;O9hyOt%|A7H2}@e3*?;Yg_)y&;Qn+K>(&}z9-|E$3Nx1
zJ8$_HPyD)%`cfSAIUoCd|4ysytAES|%Y8!JXAak{nrHRCbkkNh7c|tNp3y4}4^e}A
zPkw|}&#GPB9?3y~!|Lb?#8<~jOQWePAeFPKrof0?{=vH%0NeAdaf6D`UCj5C8lmY4
z14)6ny9am*Q3LVWzwIylTnBqr4yW_Nb>-ju+<W%_@YW|j__>R3{pufidpuHm`**);
zYXq#z)v|idY-zX=Q<4vHnw`L!I$xu#8C^_yMh@7W0}J34MmAcQc#>cQklCDb9Hd&%
zskY2$uNrDokN<>a^ptzdEa=FF>s0=9o19|{PK|=+$y29z$rD%j<^TDg`-SP?0`nv!
zA^5Y;+&El$L2tbCGr#Ne?tT8-w-`?!*8K}UdrOV3z(tenb(;rWU=-#dHOFwYN`Yp_
zuxA`&!5l2QMIECU)c`lh1Jr1hwwO6DBXR`0sGTvzr8_lfk7Zj*Y34uLW^}`ZfFOa`
z$}>9bMJDJ(3**IeaSF{e#h*4m_jmuLUueNIjxopG0`lX}8hG_<|0#eUS7#rz-{z<M
zwy}|H&&&kkNV{e10tz`mG(D>%7dI|j8r3E+BlU71-99@e(a6DN2dLVV#ab}hy(gdr
z0}%q-)$Ku-X{$J0-Zy^Ph7C!vEX$G-Nh)%gOVFkpbJBzF{?MKn(u_DT_yvFVJAS4$
zm#k5B*F-QM*_Ag32{#M#&7OH7TMU<b0xxqC3}$Ms<Uo`%k_cl81CbF7rbNvoRF-CE
zDjc(#j+=x;wh&H7$<1<)Fp+3S`4mv>-T~u7ciM&(nWq*IE~H>DO9Eto;>j914}m$e
z2{1*8ls^4``Kv!#bLfsS<6G~3JqqF%{z{+FwXQ+#?|pY7Pl4JEopKsU76v%?j#0CE
zi_(~6(G(YF5=6&J<?MqZKyxIhx^E^w1qm)uDVL3k3YKTOC&pfGV6X=2+0)5)Qaro6
zs>fc{FjJ>pe_H=1Y;-ck22gFo{={$jJ3m-SxlVUQ-rWp-;Lpfr5spO3J-T~*Hv<T2
z5f&?y7HIZ-xpW&w1gb9r5*ha-P8C|aL<Vhk3+`D+Cd%W~u46u!ow=uz?rgN^5zSgi
zunBd5(0!PT))Vyd+`F9VHz(YDLhL7=901Sa<_sN$(8gnb;d}nYMG}P*zJU{i3NC-}
z3I&SAgKn~Nv`PQsM_$@OMYZ}A?>FOeOIF9sCIcPKJsA}bUD=<zCDy?`V@*52O;fHV
zjwajD-RfEJ7kWnKoPxL{quJqP^%T1haF0(on{Ttn7&au3ipY>CEV(QKPszV0#2@$6
zjnhqlZu`E!@pJ$1ruXAEs=GaN1ZR*2KKU~*U~mm%JKEhzfY|>r>Bc_jn#RB;*bN~7
z2gIO;Ck=dYj4Ss&`e=Bds8jvtW@x3AMn_VSA`%f&Fo`}SLUQ{c&-j2fDF6|4hPjui
ztDiYlh_+G9$Xu5U5t0N-T5`f?cfm0F#+_NN1Ko|Qf8~SU{;Xm$%{JOSMsYR`=lZq=
z<4nR?U|IzM^e?=C1{DQ5(OomM!#0dI+KoZ)Yznvlt@D-C*7hh@xomnPP#yHFkl3vn
zr(!1CZNxVMY->g+bW|cv={Hs+0<ZhAoiDa|hNk)#6i$adt>xF8pr?X>dHS6=`;%w?
z>yt7`h!iO><f-n&oj3fShx=ar(o0vbUcEHRpwkxI`N;#GK4~-w4K~`g-13a?k>$9!
zlY08rnSA5Mp75M5G7-W#3Bu9ye5QJtqaQlMG0*7uMu;NVLX4SW*8`Zo?gqx?vv<uT
ztrw5EN&2;C<wIED%-24l$^d{gke7%4=qLaA5BTmmKK~Q(q&y(?OP}}E|Mh2Y&-}rc
z{g+?;sjvF*E!fc}c<~ztPS*<+HHmmP#TIO^Yy-OU_)~ZKg(t6Hzj13G;gA`iJ2W8m
zVryn;8Jg#{!?q`@mxHu5Gjm|MwTos&l-x_l{zV`D(HHM*qna;*eiNGF1cAiKd=i-M
z;IsPt!%yPuwg20Ddt!w;ef-k_c8b7z>HYWonO{2lXZ3+^$VXrKv{_tsm7lkq!`fxI
z7zF^Kp2l#yVs5$d@tfCP{?zftoxHT#W;9TQdfge>P^l;sijxNKh#fTC&0adkx~~`O
zNnV;m-O-lOj6t<K`nNv&3v;ra?*icS#?h&ZnDZB1X*V!kuS8${ro*>gEbH&-F(1@?
z>@$f8rwhX@#rgH^_$6Qe?zcU#(FVt|jMOUx^;9{4*$xh=ItnZep)}Y~>8kI2ES|<o
zuZq-2nCmd6y>rMU$Zm6K+hMv!Hi#JEJ|6PC)Q5@=SF&snOU0e@;SsSPc3S)cf9>rj
z@9TVVJfR{qg|<#NZBesj0wKKe^M3jl?j&vw1YqQ{_pFpKr;3>XuD<f)i?^#8pM{sb
z>+>h0&Y?w@e&=RbOb4cjcZLtN!>xk3hg)0U_m&Z7FWq=x@3)oq%;M$DH6N;(zD%4V
zQ)TP8baU6xI+}lfC|#WNo^vcS&3#Tk=tL`^{V%_3-dZp2@skcp1l4v2!6K+p^OSgy
z*T)=zcM#y_))r<2{yc^OGZtSyKfiWdzUX87Ghh1y*RH`00}uZGPbvaEjRzW3Pk#jF
z02M)V#MjMNp8-PC<%<`Lsf~)7v2Sf-WNA5a&vpkerhtJi0F8>T3t9@toJ<joUWq70
zw$mv&`uxrBxrP_3T=DML6hwG21oq74S3}E{zPPzJ(`E1b*lSc9RLqv@MJu_hE^b~s
zdR+VHuWqmZ+3VI)0IvK`uO2hBX~b!E!E89Kimo|_&;s{c4|7h$Gq8`-L?K(#8rf=u
z<sr{hM~e`g>KWqO8Q*M7I2}$P(F?1oU|R|n*d6=fEx&jjBzY?PaW@pejm4k3-@$2Q
z5UGrpK6vSTy|O7#wj=^NwE!n8`LCsFEN^;dp03Bf^0FrmBg;7Z<FBk}p*2gNdEC}A
z=0i?EQFH9p@xjY;C?g|yVJDiYL=F>}*CM98$Z1k;D8@wdieP-zjI5PgU5imwgk5G6
z4Xf$4&Ohxv&-;{yGoQqKGj=<kxPw`Ha)zlM7GS|0VAi#F5yKOD9JK51X5#CA&ySAo
zx^v-v;lIr{eb6tto`2|#qYFOq;uSSr+m2a*G|djPx%csbh3Fu5_P9yaRP2_iIit)F
zZEB!rj{Xdax6m}hX+s(`32MvH5sg~t>9<82Dpf!9hri<&`mXqacV}t<!00gHIAxg8
zJ}&AtU-``SdEL$FiFdUVr=MUZ*zq?+`IMvk{GYjj(kDDKKA~r`zibSU!vrhZs7d}*
z4BBLA^b8*BhNq1$rf&uuqG;wQVmq`ddQMB5k_NOK*dx9PjA6~=rH5iBv)TbdyJ(bF
zWSxK3wI}W>qwr!yM3{D_?smYid3MQeEe||WJYSaaYe(lCn9?T&DFEuf`PpCSiJ!@_
zMVvo<<<|3`Wp)g<5Y3u%T?hhl3ZNMb2N&Faa6xE?gRI+B8k&(vaMC$DcP6YM+1YB6
zMyvB~d{{$j=2%B`34zXlEj7*nIOPH3@OvKkSGP|P_+6F6R0(?v@szwOles`e<nsD!
zcgFR@#mPayyU*x9xl8hu|0l?N$??V=GlI)Que$!i&s;36Jyi`e)DSQn)If1g9x_ts
zQEJ~EFi5hCXhnAu!fvD4C3#~<+Vrs0WyNT8u_C^SP$TK+vKJ_rrpb!g{bCy8)E%Gp
z%3pl}Cl%bO?5C*{Al^O|ySlRaxSlPCi~Zux;rbkP##2(>45Xd>5rKH+dq2ALm;R~W
zbn92OTsqtuH*d6V-t*crT%4zDm)%-mvNi96Ac!)@3lA@H$$*!>4XY#QIWY#9>p_qi
zoB<LvX8Rp<IHcOE<*KeG2zPmJpb#$1nMv)AE_2kc{~wkU;(@zo;3P!oYA}dy<4i}W
z`{JgqUBBnrHs-x~x<7H!yiW{?mtS}Z9)8(xed%ML^;qrKufyZn$9KFZG8m%mIp$GR
z?QG6pmNrBUGNiGWr{njy8FSF;l1nEQ;W@YtJ;#3|q?7XBL92yGrQJ*CV78mlfexR(
zR8EyRZ*TYhymgXC&Xe%h%mQ;Y7y!<W%Q&j&xHX=*=mb4U<rxcoLIP;QeDH~P@=-td
z;Pw51$GhR^?o08j7Yk+7X~G{_(O#W9O<C;))3w>iCCMCsBkr8f+1TJdcR6WcbP!E1
zg`LhlpgG$Hwf4~1yMx-jS2M?`YPYGf7nI=Lrx=O~FYoS&2-#CAIKTQox!nzny>D&D
zW(K!gH51J+Lsr`E37*XL<JW%hiHyTxIovu2?EC+_OW@N0K|sF0FTeSd=z@bWd*gwc
z9cuv0IwdE$xL~lnkKwX?PA)hl+FLdni@CB3&8Q-sBkX{<3yuLQC!BK{V<v=$?-E5!
zrBZMdO%BETJnbMlp`Y_hiU>`JfR4t+);Vgwvz_0&a_$k`4cY&-dYYu(^pv@riC3;C
z$$S3bZ~D)KZ9o4vjK%4kp}Cf6j)}^6R|KQQNWki>X=_6V>Y`)5APjg8l0(}eSBr{0
zAwuP-_JnPNvRg%~5kzL9Q1g(Z+l^(qQA6WnCu6{g9Xfqc<Wvfgh6j$dZ0-H#+1b5s
zzBgXnAg2oWu2vSHUVRJ5`{8kGBkud3UvvD%RrtUsJ^A6qBE8H=wOIm*ZFa~f_zuOD
zp}`4FB1SP$Sxug@JxijHYP7^5s0E_35Nk+t`qs=%1O->*NOMx*2$n@fRW6W#o44=Q
zfwPv8l&RAQyh#|?=wsdEbshUz+<xI{oZrK{enh?~B0H50>-#r%jZwl8um5w8Y~#{r
z+P4qOKO$W{HklqXq0mj)HL7=%izwAaZPh!4?o&aS*rKZog?4M-xv>l;(6eQM5uKS7
zMC9bDXApCC$B3E|pc#<|izBiS?Zk&R-z@}il9Wt=FlQ?CoCdn#(kSkGUwZB$?#PEz
zHrA;F;E5C>fQLTh)8vr2_`KhGaPRT8KfC#T5551{tEk{M8{oU>gNk`~#=RUsFD-If
ztC_Q?4XPRY9JtMJ7NP>`8I0#40<En9Sko&kny$O^eKkC<ZnljX&15rT{U!e=m>1mL
zRg)#xoR(Z|>^pAXDqLLLdy(sYyx1<Ry9E%8*L;qCEno2!@AW-Kv@>7wI}W3We#L9(
z=kDWoW`QLJ)t(uMQ#V{$bj=~$cDeR_+lGm1ONW>vZ2;dD{ugGdC@b88Q|=l2Gs*n`
z`({JPQn46aWEMM6A(p{?S9`emKD@X9<}rTvCYbq#_T*5v(Z0BS=~-TR&108XoBJdm
zuz@f6{hyO}tycV;r+gqIp$H9(-mAVWIBs73>09tFsa&GDGz)C*QmqrR5(3F3<B$<>
zhJf18H8T}OD%Qi<p>Ztj4d+M<#~jyT2GWMS)j+`2r9!c)nCuuB#O6fVK}@3VBfr+i
zo&b_5nCJ#zblHZ>$hvd5dFSI#fB5#LN3H~5_k@sr_No5;U$ovO1>bnb*USTr1KAaX
z{TYwIzxvAKk0!&l^+^q`naGe3YFd2&r=p79M;TS86zMo2b~YY+*~5=KdKl5o8qMD4
ze8c_peOIXkDt+JytqugkRa2fHp>l|bn)Bvg_%$%&?o*hQsB2D*nsuVE9d!)U@z$Le
zjt^eBKj(uZgx5UK{X2g9)d_yBj}JBaGJr=8MxexzUqX(X=RZ=x;sQ+eZqyiVjGC52
z6J6fDx9VuQy2Y|x^j766^m?AlVOg(UKD)9y$PyF{A<?1m0+}L6jye4pY1$AL-tBCI
zbLEmWKEC5cTf|*u^el?o-Bl-r9L8w2AAa>YdOO~DE*^Qs*%T4m_=@T=zW9%SVhF%X
z-rZeLDRA~c!BUT|FFvC`@Ve*TV~T7H0JOw3IxGTJWJ)!U#+><8W5nLv%i(r}&mVsD
zQa^oTz5iv8J@#<UjKJav?5LVf4yoDf6#xn|pP-#y?2MJ7#W|=nGS>I?e~Jp8xShLq
znr&tV@Ecab?YjQ)dmmWa<KoiAbCU_9uzqs6dwa#7`l17=x4azy%6oMe%1@Qeef=EZ
zn9KLEZ*Fyd>C$20u?8{F-8=$q(KSY)_Z}*a;)Sw{BSJGZ9?o^;#rr?*b&ch~Xt`6(
zoltj)$^<)W%I=CZVqdzET^@wZ(P?>3$*w-cJN{GR<HeT#?slMp0R1-4$NB7nxF1L3
zS{3io$DgO&W{kt{|Ii13FMX?D_SxT~nMY>V1MGn6n||`b?|#+$uRjonvq-Y);^MeZ
zNLn$%O*G_eE=p&w6+tKEp3Zny@ZuSt*MsqdjXL0zd1+)rR83qohL(A7NLrF2QX~->
zVK0+<o;kM%jdvO#Pu1Ux28p{)83kda)-@;ly+40vzj7Y?tu6Zm0`M!%1qtlq8W86f
z{O`Y7p4abGx3#P8{}lN8gZkbF9=dW^$g?X~AAa<qOM+<@r!62x0SYcLlb!4gwBVV#
z*X?@h$tSP%{_NAwTz}>`hf#K(ko>!{sS#5i*+cM#!Z0{zVX=rxvgWvWgZYEk6nU!t
zBO)d`JT;kQ#N^2cMl9=+aNWFiqagG}%`M;u9)&tR#E;!H&aaKLhmJs?SQDoV!ejgD
z1@)l^f8u5s<2Ww1@_PTH56KW$oD>Pm(hzuORNG0RyWMmF*<Kju>$;Cf9+1nTPTF$x
z4ly{>kg-TL7&F3tJ*j@lP|sd#pR?o5hPZuy<K!Jh8(`?f)&j8R?Q^ONO(}!kkETw?
zh0*x3I?dF}_m}a?U+jkt;*xYRLh7Kz@Jn7wKK@93H&CZoT;ZN858qQm&4@&h3l?@a
z8PaT<rUG-Bc#G-4!n)YD8tdrkAFrJn_U@`K4_0L-%6DO};2uTp3pD5LHR<p|GS0s9
z+0#%op5_-ovU?Wz3>wqU10m)dMmGYS5*Ib$_1C$`CY$eS@Y8=0pJ2l*MpQTH?(qKf
zGp~QV_Z#>8$TzJ(22Qi6?mZh%J#=rlun87#&AD+qNpC`N3duEHOWg_>VYzNyj@@!Q
z%!-oS>}a++6Dpyn1#BoW1teu8hAF$~4(itT`>q-eJyX&N&N=e{`c&ODaF;#>W-9*O
z%E5T;Bm0yPrhgga8-BuXoN>(4Wz=D(Z++yceB;}m$A@13%RlycwxXjJB0T!OifXr9
zHZ=!<Oq^-XZAlU?LxM5YGB~uD?lzh2kXtmfqd{1SAf;llr8%vVW@lwb7C5z8()y2Y
zOom2tAi81k^rC4}C#T5WhEJZZlesi8ttOvaZE&Q#T?6y6=lQCUr}oI~LEyN*{fT&R
zf5T7tc%DD_=5`*|3%Rifj)$Ip5;Hh?DDFv}(#ajQw$Rh$<4_4i&uMDe%SAERP))w1
zf_0ib$fO&?3S;!rq+?`*L&HXg3oQ$A?T1a;bn8TqU`}Jf>@cS{=0y>xiM!)r=1pN3
z=br%^JZtjFqj70J#)IAvN$`+nM)$H^)ZzTz-~96z{-p;WzxkRy79d!)#NH7;-r8jB
z8cO7GLG4tdvca6Mqyr8}jw7tR$dxezg=$W2l&8-}Q9=$WYHzeKQZ&ll0xd_!)VX#l
z|LKKKhUw|uH=+4?e+}z4*poCrUHjsR6q%;_H(cK7bY<yX1w8u8eeb#_7+Jl~=rN9Q
zxOC;}+2v1pyKlYjnScJ@x73u_RnH%;<ki>Rx-&*KTC822z3x4AdB$>_12vFFkx1IL
zPX*V&j170QfRP==vN~_}IegrUc{b<L)FdHMQ-6piIo9WX@@^iy**cK|p7kweXbVqM
zMDC*IcbA3k)=Bt^+8D(iT7Mqar@ps8p;1JSnN)$+_P&qa;GL(hpMTmrM}P3@&%J+*
zXfEQ}rRClX;L@e4o3}1HjNA&|j(hmfP0Tq&R7WGWnA0qTCA(3S-CV|uA0Ocub3R@^
z+ly*%@KC;a|9LGjQ#TmUI9Y9$IQ+mZpYow#Pg7xwJ`Jg^yV^(3pwrV{m^0+=X6yIa
zYX-=z&$Q-;L4Mv}!*2i+vNea6xcJ;hw4OEn=wnat`kSx)$FE<_)%y@Su%~t*F6~d=
zr1SD>7P(?(%5mmOv%~vP)1#v0-j%ENp1L_ybWh99#K&{5s%8V*MQz)8-m{)hmvB~P
zlWeo|#*g;g?lU6TB|<MUeo~`ncnz%yp%Zi5L7S)idwc<g4`=A*aJYQ|@|D-{y7B-L
z)kC|=3m#a`uRMG;boopEQTc_-Kk?rzSoT}D4=P4<+D#uGJpbs^9nmr(SgvgiL9|X&
zz}qr)do8nLXm2!|_Li_t`I*xkqs;>aX7uhu7k9hEEH^l6R@QO&q3eLVszwtM<@SpU
zo`53fB#VjqPtoAyoln!dV18xe-m91Id+5H)m)<w*;870(BBvV!xOU?Vd4A7*SI#aU
z;xq4qpZJ7ZZ$F$}y#Kw=zHqVkzIk+w(c6CZ(52gBA~K;|fibyxwmo{r8PL_}v8ahm
z0#=68oU<<{Pob<M;&?`)bs=U?{#KsjuX`Hu`9r60uo=h$=`OU{npeUcG2c6W?Fg8$
za`WS>s8`>Ar+02^H!r=F<kg?=2ZV+qArbq{>*q%*xa1jj-}Bi&=nJm@{PXI)Z@=&U
zx_rh3anV3<8O&Gat%ji)TggZx1EzzWmBU<O4I=J1jT)d`2)idWo@hCmSzU-7LUlFP
zo7b*BaQ|RslWBN}==Ddgo6ub;yiqmBkI#t%Q#wTDW6!)ODV)QL@7Bf!&|m0V&l!|+
zx%XXde8w-<C)Bc*ouL-@S`Td%J$6*z`TT!-(ZBN8<L~=~`#$@g%SUA{b!SIOVeISq
zqfd=xghvfb|3G>2`7Nv0ahF+2$FkP|S%=+#$Z%&FrA8-`l-$SOeUFIsER~Q|V+8~*
zx8wXfYl`CT9`8n>7?T<{X2!dZA3K_-iCS*Y5sD|pEPVf$on3voIv84(>j1v~ultK%
z#vW*galo06`}xf~u~Z+SRS&=Bmwf$mKl;6Y=;gO`&+X2A;~k5m>T=&aUv=;8nG3KC
zGZj*^E81u?cGr2K!+onSCGT9Ew{jl|V^34P=5Yw_tV4x4+*WUgg@;Sg=5}~UaqISj
zXT8R?JMVb<F6uG0T{<D5)a-{i-5<VXL>MPaC0`8Qv`v5cJxCe>AU=58(C7R~ymlPt
z2q|GdEVq1i&y73#@shT~==;vT>fP>l-2b*~aqEBr)?0Z{T$GOoRT=l*ckAMKV56Hu
z1FQ|`7&QBWU3boM8PcH+dCwiK2V*RYvw9JV%pBri;;7Nf?R&!a6hV*Uao1Voy_fmK
zGglTO?!WpM`))CR_X)_=Mt3P^gkn$bOMAwPcfi~)audYiYp^-y5L~{u=;U#E=^+^M
z!OmTA<$>~jFWJwpUfGv={0=_p(I@fbfA$}K^y^}(K$g&Ap7#}}(6xux^}dVSgt|v9
z*1h|kDD9*7i&gvZGLVO}H3xZqd8}qF8j~3_p|fSU(>;RwMV(<cV*t-B7kc2n^A~P!
zPv3d#`T2VPCtZJ!0qE520|1)SM(3J?x{aCcZVcpIrev}vp4O@RU*fX~;!x`^0(|h@
z`y*H9792aq!o_FnrOWrdxy%df&wn7k;2XcK>+yU3<u^owuAOATo2t#x7;#8K7quT4
zcI_j^MK?vAElcjus*UzVT4U5R53Mf<C67zZeb?p&?98Xe)x5K9eCWaFj`yss-IuuE
zt|x9hbNi)r?28+hzwed-_uUrHr{ZoJM7xPO5#As<vXF6{ftr@BwmRIOwT#&Z2g&!_
zeEkpgtCwL3J%=e{TpWUjHO3b1*PK6GpLNd-Jo&m``q57vMU;m%M}KXLwyIh~9=GpY
z^p%5>=XG2lEMkJa;1Ea#5+2>LeeC9)*yjx3;sGOfI}XRrM}7V5o(t+VpMUusAARti
z;{z{z?B%|C{ryjW^fsP8J5(G$^27aNg?riqr{a#25pwTyj3of#EIbM;>9(x48>YUT
zBAYQ5^7tg7&--4y{&t^t>SQ=Ozrqw4*7F}_-?;P4&0DX2r*G%_kN&`&ao$w`j?oA2
zVxevwuO4)MeD0Z>hl|6KhsuQs^E_k~6=m_T4EM@)+;e|jeEjN~iv9TfnwQR^JkHjO
zJ&&JT@4wdUt9s&@*S_ZX%meE)AHTrJtMSyeXCooqy6gU!7r?KDB%M$ZqxoWjj=A5F
z%pIaJFhT9b{f1Pw$)&Zt|CWI_{w=>rH`Z$}ZAL_ZSGR1C8dxu(aL*xcz2V)PKXBhO
zdex%tkxO@b<poiwi@F?GcHO$T=fNw?><k5$gPdl2byW&=DNqnG_Nx!vzJ2ZD|5x6d
z{@9jY=Y4Cv)1J>*r^Z`%?A!CqW|Q3%2Mv}YTas-+a$q@d5+^VmAaNWdo)R0f4G)rS
z7<OPFfgKzE-~ovZ!2>0W7Ac8TOC-e}*lc$9z1?@LxlYxY_xMgLAI_=U%-@izFLgei
zU3>3$t!F*YZ-K7unm*aym`rek&W7UQZfH_pD!sq&@cxOuZx%B*?MLOt<$Fg{&ltP!
zKT&`$zW>S|a!~;+s{sNa&;XusX2>BUAoR<W1g{eBJfH{1)teSb_6tZS<NCxPfhPie
zT7{8S0WxYyk5wMD2Ayy3(Rl0p-~Sm_7?Jw97il?k?O-xx+r2qxe|mmebk;M83Q;hS
zN+|SzNDP26!8|a=nfJ^aGAdK17#T0_kAbxw-fCXlyn7_zT~-ApCX2Sszc&MJh@q?(
ziQR3FjjqYVbcqPW6)KOdwj^E1Jj<8;MXN{>JPDvByOLWE$bdZ3mxBhJbB2&7zeNPy
zp=G~rd{n2N$N}T_-B~K8s<YAU<JnNR#Ze8>kN*?$C(&DPUH!`O`seH4I)Ci-kKeGb
zy!hrj;m<zKNED+ZbETtD2Fh|y5J_a#g^Y}6AQX9zWIcHA(UU8g??i+7v^gF2!K1K;
z^D&7<X#i`S!N;0x!<hHzIkOz@N9W{#!Kov9)CuDS9AtTsmoIx*{<VAuNYCJcE(^q%
z$E9s#*Lnv8oI;iUG5}tEpZY(J`NnANSBZniAbmZj{ABdO+x+QQ*l4tS_58@}yz(zg
z8ecv9+JAZP{@#=9gCc)=eBkeXIyQ|9`oxmcBndPn5pW)QP@HQ|WCB2B9y~a?Brr2F
zz<i;rk{R!TsZgI+eK19iV&NSixY92xU0{|WF@vlmMx1<b3d=e2@*KSA0ahrzmH%m}
zlQLS$19&k)UiFR$M1S;0$PpMWkYO<cfET~(pMTtryDUZY+;NW>sbq^$o(C-o;#L%?
zAPS#)3w#ybZnv4?>v835#>k|f%ikvfSRH(D;!&jSL8rg&xn#_uK<TVtz!<Yj<A?=7
zs`F_%&RN-<u(%zIP@h&J&^Go0k<b$?V-|qkF+y0%UI30mQLUI^OZmh9;0=F)Q+YHk
zy1**q@{C|Bkq8jC-ZHK+nxkj_R0HC(-=Nt)TQsZ|hQJcY>95TP>1=&H>BD%jH6XXS
zS<K-1kLEPG`ryM30Y)K)6%Pxf_F9Yv_ulyKS%VTK3)?%7ZS6#MX_RIjfGhzQp<$kw
z20a{|PeN@Kw+q9Doy-+MWxz@Tp_d4Wm7wpWqa{oRh!G;FlZ*HNMJ{@&iNZ?l{p4l1
zoD<`UufE)O<d+R3ar!5MgSC#3kO@y9Z=BQN>9kV{=9aK2c7CU|{=t(syk@bF1e{x0
zHW0r470^!ot50*}k_Ulii6%R{Z{~XM-+td(WIz`6{P?}Kr=H%)IC=)atZ*(sj}8rS
zpPvqg!~_UWj4S5twCr|BRE>~Gf&n3SoH#K+5LP<II9UV^9$7?_<=Eqbl=sPZ|AC9L
z%tXNjQE?SAT^$*O0H{igMPobxGS(14<}2u#sV8*gyx_>BvqmGY$1}RFhb#+5#%}8B
zm9JNH+TZ)M6WS%6#JeB~nAK|b{onk)1rF#rAoEZk9DnQPi&s06TdxE<DVPRt!Pk55
z410<r^T0_Gli<lNQpL$LK~A)E=z@}&xJb*G7znUYnN@Ke<e>}Gllsvx99H{9mUsS=
zqzo76Nbt<BlrOe2gD*uwV#_iLg(`ziGb7+RAa2~Z(HHZtT6FN-E}HrXCr|$V{f1mQ
zo`=@V*`!lfHmywj(l?>l!M6Y_2;<OD;9*WI9N+#}sEU^QC2&Sa({H`^*o)8fh*{~E
z>#xaHN4IX3s;3Ns2O@_a7zbA`h;wHu>%}!1<ef1s28+R2s|pn9qC4L_>Mzn+dy+=)
zdbZl`xl9UO9EXo=p)1Sq%2h=O_(|EYlu!}a6C9xv-OH{Q2aMr25#YrS++*AD&IJ9_
zUwrD@d;YgCoxT3qFMTDSSg07%)OlkU(D%0fvEMUv*4|5rMS{Fxf#LJU;po_d1LVj6
z7$Rfn3F>$6-G1##i!7nz#;NAy)(1yC<{TI}?id5&)^kC=RNVQBhzn;qS|h!8Tvs(Y
zm^Q4`d0Z*ZbhWrkZbi!c|4hOkeZXZ+v_jDk;-iP-5p<dP=uZ|%fxk9QlATb!M}Y99
ze@Rcfc!L3M<cH1e-}vl;4R!mef8%{)>H}R8FPa7`OVtzoA<bTT575_*wUyz4Uucix
zxmk{7I3fea01yS(>9-$#_S0+DcqxRP?!R-V5LzqmeA3hmd^M~%f@H>+Gcor8*5ZU`
zC6%g(=B?a0g}|D|de%4|CxhUiT#k7jfkF{~;5GlKCSH!jV08?_Cnq5hc;wF{!g1AX
zw+wdD_$vg&KJ{1WWoJUcirataqu<PTzR`;5<m~$M@A_aZ$f~M#1_E%S#azGqYh=&k
zyyqKJhuV;@r;xGJFfA$sA&JN%zj0ImH~<c(U;gMzPjU~|PHug89JlI4i4tWftzmF}
z<R!|~cs!f53)_OJxFRo?>2)~m>24of5b&x>brxi0w`pgsUtb|x-NhLPD`oH^L$(|Y
z;RR2O@IoMYR4)U-YxkW(<{fLD9z6yR2vA}U>5yG9;{H>c|NO8$nXiRA+rM__PydUL
zy4zuVwqKN0!O>SfKY2_ww0~_cL*lNsrgm%|r)0YA6C~h)^+Z6GBShkY7&PBIdhN3r
z>fyT|QWOaz(P`;Vi8Jt%nrX(lw!Cbrh6+0sBN-UT%Av7l0%erRnQw~4!))GiX&%3K
zG4df=0g^6Ee*WTVzbKAZH^HL;$tQfZ=;jf53Em^JNYw=*tUa*dXtEerdiI46KWg0^
zm!rq_|84o@=;#00%;qOKwA<xGH2VH&YJzLG$u;3Aa4acQBZFc0wk9l^z|__VB<}1(
zBo2t&3oqUIS0B7~>2z-xIT_<ZrnZr21oM18X}A}5OuVdY<0-dZl*-t-aRgRyseu!E
z&Sl8aLc&fP03vcLiJM*IJ(q9!NBrDnAd(ot(!Yg2P$2@qXRlpG@-f<2dY!zjiSUQN
zWnVcwrbAoYc=rGFTd&@E>2Un^JW2n4^56WzZwlAmwZ-=JZgcDQ>{3;}aEIt*ZFKUK
zlI@+wRNK$?`0=nYCh#6v=dsZPW~FeJJOBz-w?*~v{JFeXPbb7!017!sUR&0T7MgqK
z8H;S=8vx@#gmJqHIFqcMIkXuluVow!BEe)7{+3_9`@sX_<lWK`x*B)lBbM4DfnYfq
zV*ZmF0SM_&yhsEL9Vm|;{sS?>AN!~D$HlK46j*$TzXj4=O1BmVb|UbX{<pvI_lpNJ
z@6^04JHt!8?bKe{y$^=*@6cKMn%};nyQ*kagE@4BPIA0*ectgipPETyGi>@v)9o+j
zx~*Z5?w4dcB$IFv^hUWWCbjp-9D!>rSjUYKBFGdHW4#cXJl2iJrfH|JPKqnNc(5$w
z+0q2MT=%=0f?u$6R+iJ1|Nr88tXy$GNDx0G$pc%S55j%hA;3#_=*CPD*ZmjYIqfEx
zuA{JEeVYE_U-&a$>m3cN%$XDBo~~q(M?d~gHRbJt*PqwXOxwBDDdcj~`Y_D)i8q%v
z_l1>VZ&b7rD^)BhW!Tq4uM%0(jr~A*%B8i<B40{-OV)}fhUA5x6_&l2ZixBE!>|-g
zIFod(fEJ>?pTjEY2yg)jTRrb^@i=_)v#Ouz2dV@@K-k4G)?g6~c#w<lAwZ;ZB}Q7Y
z+kdS4FAdAW=XTG4$OB%LfA25+zQddShapbfz>nshF*VpYp!vpR|HXwc^OB5p#>&`~
z)?mm?xyLMLLP+eS%qnh#VVN1>pgOjk+#QCNksc}1ek~IMnGK~t=3Rr#bCD%<zOmLK
zjVz<YIH0;L$e2X&yT9T-@r4pDBf$K^w}S{s%;AzBw-SL@|9E_)2?Id$w~T9y+V=Li
z7y0p+05>1n;KM^3p0!^&81mlbCwBuv1svny#?D{=t3Usbix_pY(?tUihGzE6k@KVI
z?Po(LT!%Le%TmsG2Lrh+0UV3gtp_2OR#{6787%oYu491#cW4+}BO=MH=fHp<uoj6x
zib}`I-VdbBclfz1Xff5(xHX$>*XnpkK;SO2{<zFwFMR>aI1jrRf-ffnkHS=|E{On;
z`OBd~!J;h>gCxZ7I6(N~Z@aGPf7<+`KXmhJFW*|NJ%24^f^n&WzxLPv+h48_eG)n9
zdVvtEe|t)^8xN0PHV(j%2V|%fIN%P!BcgDG49EiWj7i53nI}b@Avmy}bBPQHi2G>S
zsCfqx35koczIs`oQ#Z&>6h<wdusAVctU~b1b%xy0s;cwL-S(acks*<1jw9eQu<Zfq
zQJ+M((u9EsUmkeD8{i-|1VslPh`sW!{bhjG@9|GBstf?R0;zY25fPak|G)nApZR4^
z;63wzzyzvSzv^u+zw?^V-f<+)7@+VRNixrtJ5b$Hk}w#0MjSXWj~*QZA(<mUU>*=W
zDMOBgW5CG+L(ykjts_wR7@fwP86P-P?NK-r;=cdEcmVp<V~q$z;E-_w;%?O;ApSrs
zeCdu^-a6r?XF$pWI%`>c2LRdQ_vngo^!3-D`c@*AwKFFPz`#3BPltcy@BZwU9Y@P}
zYaxledHw6!pI$k<yFrWqkr@vXnB$zJN95pu1n{K*f`Eb%S|G;Bi)E0OEAKcuL`O`=
zj2t6xkU1tgye@2jAPAWQO9Y`{7&4}!4{53PG6olhP+YQ!5r}(-#NeXyS=CEyxoG84
zKmrjWqb+(HI}!xw7Jv{&b7(E#{pvHYmjI%*%xiB5P!M_WH|xLrOF#A-z!9KS4uQ$+
zh3|l~{OkXHsX3x|0b;<=bLN@%URqQR5S<{8MG)4W91#bL3>&bH69et700Dv=`C2-^
z(z!@v-V4brF~2l@A_GB=a~8c*t-jjI>JAr{Fj&QS0m$KMqsqz?;u&2O1xwH}L0V`I
z1jZ<k6dYP0+&I<26+z#A_Jxnvlzr(9S6;gR)hK4f8K6a@UijXh`#WF!Z6{I61{{-&
z{rGpapLOp|$XHH3)ShubT!27e&>$xyaOlu4EiR6ebmS2f+C+nA-V5mzfMP~)YY`pe
zU>G{#3~@3NA!sB(&B21VARFnvOv|Y2^4PyHLoJ(pSS>wT*@9L)4Me`Yu=J=B1lZ}+
zoVW`#h=3(uAz<+IJFb_qH)lW8{*4Wac4_CiYw*H3c-OFHK(2XG|HZ%e>em6eCn68r
zZ(RGx&$rKi_2-Os$dqrKcOEn+0>Kfj5nfhO0(xXQNc2dCq4&TXX~w7#=s{bimn3S>
zJXnpIKs3ObhFaV-=NU?0vpV)hGUF!G)mL0K0dN6O`9v4GXn0m(P`^apU3kQnP8t8C
z2p}Ntqecm09}p0{e9K?r?#sK+4g+VoN6EPeEJqF&5Ez<(KJ$Bj?(hB3n?R0{&~xGW
zXYP16ir)Dt0~&B)krgOt$wRI{Xtbccu$~OmNX~c^U_2QXdQPag<lYiGbfAoJk~lYD
z1Fx9~0+#?tto;Tt<Q^rG<PfNDy5&rRF35k6)PN-g=VBgkfsY~}!e!rx7kvT}1kVa!
zo>|Wjv}#5K#6SP5^q6~d^dnb)Q#oImC6>=L<T+aAkm?BQ$KL*hzoFkU%m_o+ICiNw
zg!A2_H(ms2xZ}<eGF>Bt%$i6$(h|vYM@EuEPL{0q4F?Op@c{!Xyp@hw&%s(GiJB>F
zv{OPiCWyJQgu-#jxQ|FP5Inr^SGpdfM`iu;f4v0jFUP(Jgex})UHC&+*}&yx`)OSm
zAQKQNkIAtC0M$99%Zp#{f69ylB&LCgEH;E#g9ip6EdzS$dw=O~nBV5IS?(Qo&wkz6
zCVcxjYrtuQ0JSqsWUY5bbKuUPmBdH`YKh!52w*vLf{+}5Rgx+xJZV7=ETKUyEk`XB
zMgYJ`aLo-vWWWSt9%jQOla(z;Tg%daRq1^q6E2HC1}ma8;)N;T;>`SgFB{{5E33Fe
z!S4|u^vWF9J~;o(EAK^1SLX8ZA*XV*o-3mz_r84cyMOhs%zwv0$Sg=?Py8;?Xy@ea
zwzuRYT1T$r1V(d_6GO*w^OSLP9<}FATLkNg@j3XAd*uazQw?wjk{Q8-3cLm`omf6d
z6(D$9OKB_zbk2o$G_Afet%Bhz<MYxvw?u<qocsh!wB5={x7<>MM8IEm#y6a#_UJfw
z9|8ir@PSR(Z)LAVM;>j!2s}}M+=F$@qpygOw@iso|IXk2v*jJl5j}7gJa)@!%ie#k
zHsB<yjS@^SttELTm_dd!Ry%HEt35Hzso`Ky3a%V+*SO`94i&N{!oos`kf8A>ythml
zN2~<KnBxq*Xb8nyZiNY3Ie;H^H>(vx^k}`u#bm^<re+t!F#>>?kbq0Z6qCS#oq9m{
z;dg2G*7;+fdq@BhM1qR3Kr(4;2n+?+!dHmW!K>f?KmLdRa97BhJR*GVBcSuvg99#%
z<w#A%8U|W&U9rGvB#29LjDz>YEjZ)>2E@s65a^U5@`zMIP&J{`O-YSI@0B1&&N<Jl
z;miPm6UK1}7WY?z&L?qKycmdoG96wX_I_3K`A4sRwpzS{f3W3*7U0Meg2(fm5ZI6Y
zKkmh^saJP?S!mB3coM>s_X_~SnF2z`6c7m$y#6)(Q~&;9Z5V=;V(r8?@#4*wnS}*;
zE*xoYwEzr-X8^2L4NsRQ{>VZ=1l$65H897ycYqBd7)!pcNph%|WZVe@PHE>9A~oI{
z<_$CfRtNR!f4OqI!6OUwvb>=Sbqtq&(A6r9<@H(G<%v+Nj~R2;ve1`ZXY_4@0HU+!
ztKrQbUgR-IPTVkW(RnXCdS$)!!f8ofY6|2R{$2LR|JjKr&a4Pt{WAG^c>4<$WSfMo
zc@}HSMb#+Bxlud^F4DRVnBk}dcqC6wa|K*^Y?we$Rb;GmhLa%*IFEn?8FWD8C1<s*
z1qTAjJtfQnalev7;Zf-OV!70X10IPXp<lik@zJGPy(S~0kqOSxl$fLAdjyO+vkJc%
zJb&#CX8AI~X~7{jOq#m3AsPg4cu<lv6#nGD>Hhq`IJarZt+>{n&}_4~^LWG27{vf>
zjXE&aLlb&L;=&t1qOidlC3=s6ut-g0JchhMu-00%1{LteLIaI(f}ya?1KCOf_X;id
z-q*7KncubO7&*MS(=P;nTyd>dVjy|8Od4I(MoR+p@~T}Ij6exXEaiYUIPpmE^7|I1
z8=vxrq72cS7652OAZsZkDJ>BbLb@Q-B{uBUe<FYWUoVgcoW;-nu60^{=jooQi8Fz<
z1t15Z)|?y5Iu;E|S$ZzLa-!i1b(jLHnX{TR^2FNAd7hN9MU>tu?!7R=YR14hZxr{=
zc%jid;IB?UaD2IOUf89?gRn}8T$F}z!AN=p#=1a*tO}>6GlL`u30YbNK<5ZZzwet4
z(koBiRt!mUE)-cHMhu0lNJ=Nk2@BGo(yYeLi~lhEga4*L07*~3Y5lA>Js30;Bh*q-
zi(7`4g*Bn)%xN3@Cg2Ho(uLZSbVLY=1!#=yde1Eez?k)lF~b>oEd@!i3N1Ru{1U|z
zhEyttKPlf`LTgr8dRnbdxrmG6MPqO=6$e;}2TMT!kZ7Q51wlccH6H7D+7J?Z`JYo5
ze5Uu+CLwQ2WJu@`tT&Bk+!Ru12!yR^;jJ@C<*)qBzxD_JVZmAI<Lw8w;rRAXbL}N%
zNDdTO^rCT&se;u~a>tmdB5t^M$^!bDmjU-&k>jfIAO%*wt08WjW5A+VG+=$>l%s$(
zOvL0ZA}T-({&4awyFB=mT^ar^7D}%8imT@Mg0;JJ1Q9&?jfD4QJSX8<2au<~=Lk>@
z4k+Gy<#cX|d8kz}Ux7~@2M;a~M%u;!hBfnO16N*BzxY@F<Ucpg8UE59&^X@rjJRjx
zFz~W8E+Sn!7Fbm_z(8`#z?IapeBncsjn8;x8|mv>F%zPe+{`JgE2fQ9ffov;69%Fs
z1&SGwXflX8G57`bhFKcFRyVhQ<o&pq5H2^L;KkVwv|2q(1mN*X01+hyEl7@>soNeA
zwoK)N=PrM>ue`-TC|OV~IRJ*p4W%AHGJ)Lt#MM#+I`zHuul?eG@nzyRXm`$C-Ob<p
zQb;ytRUw2x7V|o#5)_4+TP0NOJ&zDv%#6q@&alyO(*!7;L#&)(C5EN<6;6_LR*6NV
zm@?efCQt;G^PXYHRti+ZU%mXxx&Nf^TgtxpNl?gN;KMFtV1!RfM?iQz^qPy(BYADO
z3XhQxUizM6>8I*R3j&G4^B|#F=^Ee>DBBpsOz5D}p37P=2^dJ(<$wMk{N%5Ig!G4h
z+536$){i@qAma)mCynzatVC_NbV3?a19Mb5Uqu2qGT_VxI%%|wsWvksy)IY;0SDeu
zwPdWPm~x5&o-)P*5gQ9MuRIV??RWn86JllZft4ce=>mX)uo46ru3)W8VQj@c17KI9
zP;@cbymqM*cRP`;00Ex=hLe|Hem`LXg!KsMJTwkV$x(Y0%+?}A!&MkYVQi})a&C6z
zxj$omRD>3u$q<XE{&u3R7PV`Y3$2TySRe$-8Y(Jlio$Ud$7`u#X?HCk;E@dDkVjHs
z;-w~vTS5ue3WMb8?zR9%RZEt{0|LN^DPA$~3TrJ@jqE31`RdR=>V7V)AjnoS2`&Gk
zPqqr+N*-n&dV(BZU|6FIjb-XhM@anDFH`?DwZDcrT0jQ1VazyenC2RE#DGn>WOK3R
zQ4=-D8A*Pl{EL70m2W5`Ui!NAM!oSl-k8WGM8@{4OPOy}gc#;>CNifW0tcPB&;XQx
z<Y<{vHn8nDlng4WRTZG`Pb7N5zknJNZvi+!g9t)ITBYR67=+|eYxCKSQFQ@;!_}rC
zBwX&?!Ieb>mUdUpe!wz8FFn&J;31b*n+ikvk0$`o`qTet@%Yc3?77<7l2=B1)mz8W
z-n7NI4baElxd5pnk@>(ei6IjJdad}6|9NsGo2_>>_;m3^6mTwi6@?J+HuJRzxQYpw
z;?f&twMjkasEt%arE6C%WKugv5hosTsRD^Ew$4eRMaUWysxT}GV$eXwYu6yx4w|6t
zZ1~EH|JOaXvU~zAqrhx=H*i{+qTNMrOc!+%5pIv99r1R9wJbc=_MQg@gJMok@BGtp
z%xtb|82GUJqz-44fD;|70}2fp8De82S2h^O#^s(jc=P=~^^co3Rr1(-T66v%{|M(G
z1*>E)(3JSpsnT<gLOS3~s?v~QMxgMHld`t4g=DBlgNhpw8s3C@Dnkd73#ViiXx$fH
zRIY)5$Sfo}M^QNQx#o_wupwG53&9JMgkN5X<qDM5V4N<HVJrVouvH3^Gi7YBsHG1?
z4FK@$2S$GWWDWB&QGpKDjEV~&jZQSwM9MdqRR&nYJZQ`WGsd*Y2`66t=AZvt&4<lr
zJ~lLq4(DyIa_k$NIpt$ryNNVb637L_+H%jP<UM)CI8~ZNQ@e&skR+SfIfIM>5d_S!
zP^5`RmP89h4R_2s;d2+*c<vnV#3&b%a3zrND`G3I+Pzgk@Iv}|G$+59i4ksGBE$wL
zU^>>qY<=GX19<g4-+SRdbr+G`k`?2*jV!L$WuO!j4vBcKv4x)7K$Dco%;r7ERS|gc
z?Aw3&Z{{DSS?=pj`Q8uLDyt`=X=1{LGaBfMErMD|XAFdr*Oa=tnYqCF#zGYbj<wNC
zOO*qihYmd<8tDYrtfd)C84Ba2<}B1cE^|Vxh2>l+R)+*Awhn5Bj9&mFxL+Eh7<j_f
zR@-IEgcpX|#{sC~p4WjD*;o``_dv*B{{{cK`YbCLpBlxgfEEVwBy!0@EEL(etbEsy
z7k(m|Qm0XkUc0DvSgRM_`HO!C@4Wu21zr8y^g(cz3WOL#qmgl4TPw)ADl1b~X%%B-
zYcA48#97f`NG7rxtDv!{If2VETL$GKwL)2M9Vyafz_18;0WmsZq%=`&d8SHYbq(P7
z_1m8O64r?;<See<3jR^6w8~Zz0>Ni%9YK?3Lr{VXQR^_<dHkg-h=1fg5opVxadnc-
zsya}?Fx4HdlYq3i8dqmMu%>4sTh*l?I&&coYH-nK-uSb>=)!OIe(H~1KigxWEyG%d
zrS3w`Q-WY2qaoLEdoIGFV5Bq2p<&S^R&y0Ktfop4Q@&p`+|@1N8*IBd|IwLVSZa*5
z4lJ{VYfiZ)lG;+27oj0j|E@=f+L*)A4ayiUcNqf!NAiHQtPhsCeYT8EiZ7@NOF^|a
zxpg4UzgPBu{3m`Y{o?#fZ?+)Jf<z+oI*KCcM9@)^vrq*APmE8zQY;B&1f6VCB)qL!
z{UET!Hr~DS^J`zclU>{KUyr1xB;{5xb`X0RC?chJt4@&ehy+qAb9P#jk$jQpdV!uX
z-*keCbLP7VkoJ;sD-tfPB#Nd2h?J#pQ37(VyzO{mP*bHw9tdk%A6JhCMVtX6Vo1xU
z005W+SQ4}SQaA0gh~`RI<z$+-OjC`MPMRrV+rNzkZ#95<B7~s0(j0lU2#i_J73kc>
zMI{+`i@>4OMj8n%S64N66#mF}hJ&sC@Vno9=Z&Ac85BjRrjpNn8HL7&m79A_noK&+
zL*tB=3)zVAdaf{>)uvyGfR%wHqf~8LwVY_q1^=ltra*gZtqzqHi$<`n?f_$9D7|Hj
z6~)~{PrzD2Odo+^y(i+Ft)fwE<(vQofNVJ?r7!F<ALvHApb8hdNk90o&cFAK*WKQm
z<3Vi+I7L2(rYL8Yhvk%UV&p~%KhJI9u!>nlB|<SuiYn%$bz!6`MNGT+&aeFD8zX)6
zIjF#J2$~szQ)S3zz9z@pwecdD049|+0AU<KM^lhR8IOEOY06q*Cey^NC&JpkC_U7y
zLGA?t2EJ$0xe)-NDXe8}JBzleTYyMfhiT9}nvjtPhe9y4t1Y<90f6Thz(|izWo{0J
zF^FB4v2nadzR1hXKk@!Ef9XwbJ#rJXz_-h)l+F($FseR_M4>U`Wz)DeBO$S3!Wj?F
z<wB{oQkk{$H_o^<khjnF*11h$v`uCUCrXI>{w$mDh0S~|TToRyr;6G|H<!#+shKx%
zqJ6i+W+rrD*oZ{oB5U{$3(uu*qO8#jTF#?wIAO08K@}z};Ib9ndB_}q%uvi=rTHy0
zRL&4E#+EB^Jt8ilt$;wte!pU+VL`Wy>Q)&h?$-(cX#G#WGzo{;Bh8z%6jHkh0~9ih
z2~$ptgD>Z{%pfGoP~^4_95!}7&A=zc!=&k{3i;VNP^;0;6-*9sLIyn*cs^@Js-4R%
zW59}*2!&HTkEK~4b(L4Lt12IGMhsYOOu=ThXi6uJd(zgNkSWO{G|CtwoHBG4h&-w?
zsjxXGWR7YBX_7C!_3UEnkn@0&d%9o^6JE51n=$!8UInZhs>oNni>XJX%`*2t!ZU7}
zQM%?{pC?vmAI=~e^@+JMz{w!XVMCWjL@ETY<}9&w9NcZY-sQkdWV~pe`^fuAbo<=q
zVLXZb;aYDvOh@bErg8PzEQp;kWl)Ft*0kG#2@mI$P2^f}UbN*gvfjd6sn_2xI-Rqm
z@ghb8u}kJofLq8ItW#1~E=WxnCyu+)<O@UrRN2^Hxy8|3j7wInFKFPL`Q@|09;Iup
zoz@LjtuYo+drtGcz#;&A@hu&%H?jz;n$qz&<S80a@}Y2(;GyXa#t|1H-BMjrdSR*X
zUht?&W6@mFo#fcNxeUgBw>^QnfBw!3RB()0d+x3cFy5X`S&OL#PeT^B&QOcR4Dy5p
zbLnudvD|9SMy`EUaD7l!+#fS$Cy^MCHz9f6IwiM=5@g0v+W3Z=S)^s7^DHfwl)y4V
zyma6nK~Yyx079@1k>PT?p<ihOH^M4OwX0fdPE_$)@9UnB@RhIney?Ochc-=djdTKx
zw`yHow@{(>jSL+##b6@R`9vwKO`p=SDw9guytHTCK}A`Wzw%Y0dGO9lT`6ZZIAG_Y
z+&<{^y-ztmlQ!f{TvTN#(_-qQ%~Nb-Y3d1;DDoM_UMJING`V95cEk*9wP;#w8b`$?
z+dA_JmX51wh*_KG)_A#ERO5@5FdaHvl7%0+qnAnf6(qvD<vtgWT;?t)@>bH6VWrt5
zc-sR2fAW9#H}laN4oWu0Mo#oxY|R=*Y&M0R`O)`%>#`HmR&M5#qY!!ERC_TlR6YsF
zGlYOi#Y9G*`PPEU&GYxKn7%DGiuEZJ3kft1l6ll%XU;-$ohpigtTcJYZ+e5p;5Z=f
z`oVhZBW}wRnE3VcsW=doid41jqNz2qFwGI6=%4oItnu=+7hAt`>MMqV993n;Azho8
z)hP6$NA~cEFuL3ji!k`iXxN>29B+?w7#yFZI0gVn2M0bH-x7nd>>ph^it<SuPMtUm
z(hyYX*00aog~Utc9C+0WS-w4m#yL#AnXL~YiGVd_Ba>bKxvzjN<vUN$f^mD?v!`K<
zy3dut*iss@sFs||pk2@=Yq*<<HijaqWX5<U>u#DC-JON$et20};pUNJ=en#-a@-1I
z;)c(Wa#^)_+l&)?L~~2T1bA16OsmxDe3c&fWLEHrQF8TYTz%H%YB;E>ia1avD71hG
zkCz(M2NM{s`MdIA#N)Q`9e;AI9XBRegqiNLozR}N(t>)fw8>>$@I373K?60KNhH>0
zdT^?yy&w6l8mdbN*gg|sGxK2(HonBVPV;eBo2J%}JsiyET_4~(Pb4#xM)sx=EOa{F
zHpdD#j%Tx0H6J8w!M~vVB7Fe$I#=C}SZk%b*HlC-3zQI@06Wk&Z!QhPOz_xqQK7Vi
z7x0h^&hYXxgXKqdb9XAkbU@hit!~yz)A#QYGW-etW%W{OtE4B}S-M%MfGL-?TJ2Wq
zLfuUljkQ!tw^j0TJ!*n8w9I##*v_c$FQ<}E<BnP<Cc~p6u-u%!Vw7OPmZMU&gSk%8
zMchQX9W+HXQy^v!fDLVk_1LGiUkg(7z*-}UB{#%UE^>jn$oP{;N;g7eo-7r!K^<2q
zH6#HO2I!WIBeL2=votP)EYIB!(q-WXdLaM+6oZ)0>NV;mjl^P+sNfy}5PSYRE|rsx
zRlUY^!X}x%2qSH+HSfMOAGDciYi?N)<;RIPFm?fJf{F2?c#)m9whrS-(GBB8s53iy
z{k?|f@t(i5kZvG(DLk!(2ieRMnS2oE)1b@;$|M<-<t*z-MChQ$mNQ9j+)b(wCkWmq
zTl}eZw>P#Onzj_1PkW@J{ba6EWF=)ZYvCO2w~!O_#Do%p&D*kZGIWpDOkb>;`T^EB
zc6K0$s|7n!)A#R}jQrMrssTa#FaP(!#<c_W*kkRmipFZbJ;^6$r78Era;PUb`#2b1
z#rr*)OCKHe&iWz`!T@=|@AW6DREO<>cE$!&Pq*(*z^6^~bk){t4|v3NB#rElZnv|<
zPOBiu)*ysz%|x56R>U@O;dH91B+C}QT}8d2N@NS9;{SEjxX8|+j2eYzk%ebYimFT@
z=etdrNA2VD_mFw?z`10MGZP8jSRo$i<#3T9cw{AuWe<<{r~KB3M^joK9Ah<n&j3Q}
zfBZ+2m$x6=4rSpRUzI_5#sJC(6-Cp^C7bQkZ_Elc2{Eoy>d#F_Pt~@N*<?#@o~@T@
zcj!h>M&7oDo7p?)r5s)l<3eXnwM?1&xO5o?3Vkj5=Rlqi%s6f&7dz*}+3BT<pOWDw
z7_N<VoCiT5&nhw0t~HH>Xc@r(8JQ}^Dzr1(so!JNHrUrhh|B(mIT7VHQVXqRP*)e^
z!qqGAVqzLa&`Yn3W!`Peoh>uteKoXzNZX4E?Ibj>QZ+Z*Fb9(PdFE~;;(od|m^7VK
z<C#8-Sq76(j1-@;#6RS+>}e_ZLC{_3%pD1>7ioPho#T9KeCM+tEY?&oZ~5z^ia%t!
zjU_9IO?*0YYN&#&4i9|mr0Bq=K6c)wBYmxGGS|wpbR;{1zZj-A%@bWzjpat$$_$gP
z*K>e;9@o>U8J>K^kbwuzWyn1$C}k8X;Z^MI$~3cj6aWAjbQKq$N?TPUHsVMuAflY!
z+9yQ#(T#uSeyV|7P;g`hezA+t*<zb+2G?Q+SrZhKylk)EnD@f^tZv#6M}e%<AQO!s
zm!?$fusxM(be_AY{XPK0^DB(0dfc4G(=&C*J(CCFFp~4_l&4*XDqQf|N(u@M(X``&
zkjv3r@H1N2`fNUDh1gisy&P-9ld5ftNS#lUDlYkUU28$OC8EIu0D&7MAi>G;0A`uQ
zs0ErT{6QlT005}Imz<TXVvSl{KAvWGcEl7<2m#oSeQblNj$Gn3+bz@U?X#`WMar?Q
zbOxreCUJS{{FtUyv_6=3NBgY?eM0$sJXu?V+q-R^AGB~$T!}{T!oMq_SerBho1Jz`
zRW91BZ?Gw^=;+De^j61jj7t}v#_`CmCDW{ArGC(zYl=yWPN$RwH_6Pj`HK_1XdU}j
zA`<~wCO4CmS|=&6=@S4V;C1W$yIc|X1e^yXz<D7m&!zAx8D7kaSBIiXH`+oby!Q-1
z(5{misyN8!<@)bbfGGalzrKC-<@>_pC>WZDH%7XecPj&8!Dlcb%U#5L$Fp+NGwAG3
z+w2TvGmWyvoUe~6t<<-f?=&JvBgvFC4+ypv!)rc;K^66TNg-QlbnZ{k+kS(A>eSwD
zW@kax%Uy4t*gyswG>g(p$dP;BZS!u(MW??|duy#uW#p)g3nmv#T2TSIqKP3Ycv%81
zx4e4t9D$MMne&dj$3DJr5ie)s(WOvWjuzcbmDJ~fv?7lRnW@YB1OOmfxBOG*+eg#}
z)(Rg#ibQrasp?OiOAVRzan70t{S%oAT@UgzlN8fytjudw4kNF2{o11S{8~Q2xU6a%
z%jaJG-P%9Mjz`;U9v~hC5y^46{?zaa6)+3Z*d8X*&C^7Oa%ZnldADUA0^f1+5VP3=
zu?{0sm+V~eMQ1HqfDc(xLA!rrXQSaNGSv`O?yY?20AUH8cI1$_t#gtbZ13?LKKVh6
zBT|ebgQmPZU8IAWXW`~komsnaA6@;!e>(x$ocd>XpM0vzsxeeP%j0A)G`r<-EA5|>
zoycl*V<%%esOBdk7|H%_lyn}y?pf9da5|U!Q6cl_-b~sD#P1#51hDE){&)4mW+o4-
z@z`j`&qkZRYklW(F)3Q>z3v7zo1G|+is))@bSP(Yq0Y)<sepP?;h@^!KT*RLSCm5U
zaFRXd*WwXxB^eHQc&!Dv7FKKDJCU4`^T3!?oTDv?82aHKNWV1)5UYq7wm9dKA#<+s
zuXW?Ie&Th`Xpx?ieA_qQJMusD;=h8=Nprj*uSk?I2w<$y=lj9niW}|v0aV$gaIhyj
z(s>Bl4J6}<r;TR1Hn6&kxT0WKw4<=72Zau*>uc|Xf8>{bQovkv)MowC)Sago@VKbK
z=i+RA;%5aGb`;&81X?GL->O(r6rI}l^X;J;@U?Mwancjs2^-eYk+5~!?m;sTdv`61
z$HOgqz>0)d%mQ)C5P`rmK;+zMC+jD=wIq?j%rDW?#K{@u$p>6$r~EKqGhvy<dD?1B
zz>kT2{sR|UW3WBel=qk@^XO7N%%s>n7TcT2sMs+3Yt?06a29nNhVINojy({5oMd3*
zom+UCZ~1#&**or260^=TfAsMZtH<|5DL%S9=8t{Y@&`d!L`_!iilAq0(jXhyIGUtS
z3?KS_vQ=GqG!*RHmb_8D=#7T#NmRD6XHQvD_BH!DjAdpp!!Tp)MY2?i!th3sHS1s|
zBq6fIFtUtoY=f~)X6!z{?>pyr&iB{zoO55-b3f<)>v`_`T=)GuQ)2@nk;<)$7Ark=
zh4s6EQ~*r^lMgvqSXiz+HP+X&DcM9!#7n28OsDm{oW+ZP|1KMwI%?u*+{}E@9i!w^
z0q<?|U7BmwvMOg~vx#LUUzB^nTioVjk;JHc7<n#&bA9T?X#r(m;Al+{1_^6qhYz`3
z%n0CCy(lR9qsLZCkm9pZem`_rMbfz}po<UHJHRIqs{XmWtJc^q+$2sspw08FM!hV;
za^qbWy>ecOz_Ku(Dcq35z?%tOJ-38&Yw&s>A7J>VQZM~?aSh>;pZfI+wWl2lwhNj&
zcSV1LRVTBr;<6F{B1)XZu@VPi1V~qrPtdUOv~2i~gi!t5Vj=CvbN^<HmI)$LK?M>C
zfj{qroH4o<qX1v9n$&RPvCb{O?%}R+@SHRqzW1?y^tfzoJ;tRcT+8`Uof)Kzqb#tG
zn`p=;Z?aH#+a%zQ;EpKomSe@k{5;v+VdrNM;B?iB#9E4qTiNxOY7v_8zcHK6DFH<7
zks#H2Dc~K_U|^dZGn-b3Xl`Y1^PbZ{YM7e$+|Lurd-HtDk%;I_Ej8@+2JFDEKZRMt
z@w4=dpb)&pStRRyWKXHn)7=toKAPkwnKFZ~U(U)d{&OEvw@?|7HnZt0o7gTq-@E#S
z6BM6&I374}4uDJ=>55pqal`qJ&AzzU6^_o9-`0Qk-c280$5X$YNw0s;E$MqM;n|YX
zkL(RA4E|$;(@3EVU&qnT)zoAXLNyR|d|0xg?9Xjz^et&;`l!NQG>?32w}hzBT?BV3
zm`av5N*UH54_Q{$yy*JDIs>-B3C&JqzXS`JwBF__q@<aFO2zau%LMa;K3tahG>zeW
zp>8wB`HD;H+lM+x|05KUGlAnV3+EXxtt74cV~=t72M~l}ahnIfJw5WX&N)k7XKjw?
zNdAKIEE0uR*x6ypVDAm3&+0d12=MC|`Q^^P-mDU7KKzp}WJ>ilQT#5Qg(UHU^%7)`
z6FFiR2u10^;8PW#a1!*Lw0+_Y!OC<sYXguLdTt%Gv9&YNx9J;&6aB&|m{o=-5ChUU
zG=r^kd-JZ^x~poC<3rB}=v=xs;pZI=IsPtVYF0QG3a5HEl6emk7u9ntWzrmM!)81#
z3EtHIKH%YmP!kskr~BNH2$rDfvAvX7?T$D>rB^fiUthT0lLC76<K@z=h)?-{6-jqZ
zNL75AH35bO2}ZBn=zp<1%Fs7$9*eYk8yTnfAb19_Jaiuu)nF+cExz4SW8DnYS>uiT
z{7&+~fB*Z=!G-4=g?w*tx02QZF?LF-kI?Ph!;Gxm&&YYkagqRM)^go4gUj*T)xW_q
z*J=0LzSZswIS)N#<GHd@a1}TH+zu4v8YX>g<7D{`$%UTE^mx2KY>*X!e0+mjGh@bC
zUyosUosGk+lQm1hb2j%Z%CfHbz)0NPVKF7tIo#;(y$}xQ2gp}pzlJhN`GGTWiK2<4
zNQG-YErS6Bca$GPe{%IXxZSiWn?HhURFlJ!^X8vkjmVH6tKQG&tS6R3XnEDOArk$r
z8nW!i2zsHnKXN_&><v@Y;ddje*2bv1(Jvsx5T8YK7HNXWUW}#S9!RY!*cBxkQNDyV
z<#Fx>%7v5+XljDAUED01(LKBAii1#^xb}4H%2=tZeGbpkYR67yz@c?v$k`s4(=+}8
zu#Ccc-eHy*+Q5^GYE7a~<KA<nevHv%kCVQJ+n=E4*=<+S6W><XK}58`eNVR-$INWa
z=EFPw4)Vo4u6e@e+8jF_{X&JdYPF<4-BP?1KjjJ4cZ}_D2v~nET{4L7BWCUt73U^T
zwW-WcF#x8`!m4xyL-{sK5d-p(0zk006Zh0H@$j$;sh{<)pXq9Ms)H~9Tc<vj27=jz
z`B+O4d@CP%&9&y!!D|R~!K*;my43&+LTDg6-J*^wEN+`Y6)xj{!io+4^!jUdu3DSf
zlu9Bp{w}MHvp_m4);H-g%uWI-Oxk%Z8xwlntuftPQ(ou9e<jM}&o;a9mKEI51hxII
z6!*8A7ha2O*J+MCTkM5A(N?&XGqzyR-6HRFF?1|^MZM8V4sZ`wN{eM?jCDJex(y%Q
zn>MGm)a64?JC#(mL(9DSBvT$c3y+@Dv<@)^2JDL~c(f|sg3DT`mS&uhcJck%%flRl
zwx0xL4JaXup-%Xhy>v?E%|`kq2Ha;nB65%UWk%;kY(usfpgg2^@Xb$K$xhL8|6H(V
zjYEEIW}ms|?6>#RR`;z78HWEDyVdxG=<Np9zC}&Vv1L6*oJxV|?!gcgzwCvIw3@?3
zDzzUaZ=69B|KV#aU2ecX`6XI?XuDB#v4{kDlgqc%w`pVcaq4M(cS-Dh9vU~`2|3EY
zTDKr&E=2YDihG^$HCJ1`+_l8-OX;G3VCpCJxeU9I%PIaK7PpxIm{0s(k;c*h)$T3J
z#-{0>UQSbuliVV<iRYhn_9^*8pTkhTf6uG4yG5jUx;ousUiDOK7?@7+yn9<H#60aS
zv)?0}XSUryOPLJK2&Ar2!x<tlOjr<hBO12VwmcrOlr-uxq&dMv<`gwx%9Laa#3uw)
zm5{?#&#thCtSZ0N+kEj1LXfi+*debU8{y<ii|i{00<Qbd7mSW4br^^9tGZbsMKbJv
zvTwN)0c|8EEY0NYnovJUt)~8$C(}L8HyjHD1G@KbP6<KpZ_VQSiq_g+pDg=OW)35C
zbOV0`LzU|_DvRJ%8D<|T0mNOk7^Sig`pVg2?VWLOdFtaYv#o*#_wI0Z0>!v{c`R2|
z`n_UhSXtw)Iz=|@-l0XW^Z=TXs|%`38~SjEb+5&!)51r30Y>8T-}F>^e{&}0t7fcm
z>fb~_{ZJ!SXs#|Vt?m_ezDVd#L~$kr#*S;%1uSnXYpaAGZSN}cW!w~IL_!Dne8t;#
z4u&G}K=_It+0%ndiJZr*SqwC~dNuRC9gmg)rpL`zXo0b{Cb0t9bx80=zacpHM~<uK
z*&me%orv7@;pe_X9Iv&ooxg={amfvn6qT$tvW{^oQqsB}LQOqtr-6lrg>}ypXRJbL
z2khiBwa;N&(v1_tV+Co?tOJ6n67To~dB#o}M+{WAMU}GOZ(M=%JS^m%2-gtJiPDCM
zb{a>M8lxOM4tv=uvnXMU@3hEzW9J{|(5m<0vlHklZBC8af*6wZy#s7Hze%xSsQh%s
zn<CN`<U)&*IsUO~fYtu#W9QmF0^h%vp8Wou>=d#Q^lzGxl!UHMo_ztw*o)8D<SP7L
zfYxT`GqFD;fy==lA?TL`Pa>oB9oxsJPnaxI@l5@Bs&0;(rfTzxH^D{*a!?vs#|gc%
zO?h{u;yg@GTpV&(+A1x2xmN#-4<biuXE%PDd}tR;PH@OcL%IWbsTXrwe|iQx%#Hjz
zvfA^_hdUjgSqU~@G9`>Qh20%C+i*V`B3b^J>n2^TWEMPW9JOd4N1s_fCKo7T?)2ej
zvGxAM89#>-ld40V)cEt$Tmr{JvTV6DR~>1Q#g*@Mq(hCh5603Px2lGn>^sze9N4xO
zaG2n>1Ed^nivJY&OwFo>mu-2_<&x2vO0l(}7VLd#zv!6ak=-yNCX}Y&aE#jcF&?4Z
zB*PbCj&B?Ms2R`ZsHgJtz?kzNd1B;Ax8Kf8g!JxW?%y&Z$O&rf7}73<p=QFnUVQ&8
zo!+hktEB{aFLDHISH6(Pvj?z`rmwo=e$|j9a=Hz9Bm~#Lr-*JuPHrWJ*Y%^=Z#>G*
zT-qOUa4HIY)Jxhm-$}C9ob#?&hD@Fxzm`O+A<tVi=J?h^_r+$X-Fs@EdZtY}a6cQB
z$d{VV&TMNK+L0)H@H8DDf2}7+^6d5SL<6Fljgxk+4oA%ZPJ|)_n_Oj+lP=~=Nmsj=
zzSo;_h8ZF;c$;hzxRL&<k@YWs8=;6%e-xt;!r-|U|HqZrd_}_Q_Czq*&<rx}IFpK&
zvAGOYqOXk8+HX(44aJVnd;K*iE<0YkrZp8;LB+HX^Hg7+H6WlU&tiX}iJ^*ncR--F
zXdl9#AOWGuk%5Md4b@bmRxXW2>IXAO<TRx9<jfLR#nLsgm$6A_H>Sc%qb0)+X#Qv4
zS>H~g+zr|kVYw10RC6_Onmc*VX7QoC$f*J&(E0N8iQKSnyUZeic-FAJBK^L~^n;zi
zhKIhsf>6lh-zxf%knw~uQuadOaqqc>v(AM2u&7Albb#Q_3m1(kLEm;g7-xE(5N>MP
z{~NI>-|FPnK}%zEg#uVMyWN+4$16lgt8y{1@&xc<sV%rX*<H}Q4?Au)WZdUGQd`^q
zn`@VkD;+H1YtDvas3&ewn&!Cccsl>;i>rMguQPwDv*QoxYz`Lg-YOAfmo#Qy>!X{o
z#F!{v`QQLRJu2^quJgM35H}@c9e`zq#Yfm;`OkaPcqA5Y^}3kG_;cvvD{trkkj;BT
zORSY%Wb|9AakxT}{@A2tGjE0E@LDt<&X(We`KpZ&!~?AxT@W97)r4{?!W5PKnCUaD
z@p}Iv4NVbb&mpaVGb{TfF}bo_PTq8fJISQ}QHacT^EK`BJUcG&D=Dn(vtL9-==rKk
zFM+f#Ii`dFqO#VlW&W}E4JnCt))#C;oe~~&T3&k3AnouIbxFJTQhocHi85@RA1{3f
zFR$qxbwB;~bU7`=yVllb-k3Gp3nF3mxBGra_bv2|Ku698{hBb@pz2+$-Mq`1Fo-Kb
zTU&-mbrwC?mx}VegxfAudDT&mn=iD(@u#<=lqWKqtA0;hc)AD;{~%t*bNm*<t>QY}
zR~(%9@_Fr)3KqsoX(B&sjlQm0xj@*LxtxKWawnTbmYy@&s9L<s@xw8tt2~<feQjvI
zxs?;&>Y4tU)l@*aLPx}>c=t%7o82jA{<*o`FELe}mDC}N{nb@G%#(z_om!n%QuPr&
zjq*SGtz6>Qr;#XdIF@IwLpW}fX$_g86kz5^AVw@zOTlb@a*Pk8idW4wm)hw()xAo}
zkByatM-IH=GQ<=WS$x{U`W%`4u6H`sQyS4a6SF*J+Qa%UKm34_b066kwu^Sf412y0
zRL;>iF?t?$SduJ}f8%)w1Sgx%PdFU%s~YVi)%T*qYMp;(Jem^S7Hyj1M15ajaOh?m
ztwf#2urR+zFm_*_=W1#??*E5x61@ZxOW47KR!8OJd-(uxtgy~-MUqpHOmh1#CCSa#
z=en`(KWlLUkJs`RN1}wHqG-2uj~qFA>M`ueu@?q%T0BE=Ev-6N3GG;F_W6Y%;Gf8E
z<7<T$DWO3Bv}Dha@A=faIYCNWJxYLpPHb^N5;;Wg5k`Yy6f?|{_mq0oehGGV?5Nho
zb<aOnSmJ&A_OVTUEnJEy<-i12i#(5GUq!Vt_si!q?0tfDD<$ps2Fyh7_y$*Z?ppVW
z+rJcMXohX{NXTgBRO7u{l(=~vs`9>B09}w@TQ3;0cU7Q)1a4frfXy`kosvPBrIx(+
z#Sw)y)q3;urXT!lJBn9EhhhYw<nae`22GE?8{wIDf^-wNnm04ZZ(7UVH|(xl`{6ZP
zV4%i~_}Ld>`}OHEJIKa9gDtV*utyI@+@DPO4gZEUJAh(_HReA|Bz|j=Y!P0I9L<p`
zeN-C`=ztc}Yc2pHmAYxORAr>cO-b!LKK3-&A4B@O!*BJj<HW;aDUbaj*XYW&#XWa=
zw_X-GXGe1%ksx(V5O}=q+!GUFYVz=$jT9r1-F{t_>R`i0)bwWki|sjO_U7koymE~0
zaE^KNBh*1ji-PUmk;BQ55&?6fi5FB?kM$U8Ck9AWYAV_G$28hUS&n8oVHcZbKBQAZ
zEV>Fw<ki>MBuVR~JGe03x1|{I9I?D7zL3tRzjP#RclR78JM^RokM{Q}GA6LQH8MZr
z{xo_K1~*3Nlr}$m@ac@iAmjsvHvB7!^H#queJX=~A;-EspNpCo=#ZwpJfYYC8>8QR
zFs1bI^S+*<pJ-I;E2}dqY&Y-4O8z-~qCVrP3&SoG<}1AO#>AjZr`!H9)%lGm|7uzD
z(NMq|Hv7PqF~_JMj=u2&-QX*#=PgO7eE7CSkuI^X$V|s}7QuIcmr5=iK7fs_j4$lc
zjp8XqS6-`=?L8^sWGUZHdXUw2NoZ`+%}gQOz)y#4VYq{d_rBI-|6%j*PTz^6Ki}nU
zqplLr-4%5|34McO`0?$dzA5N>ed>U24;f4s^!bNgoIr{SD$|ulHIxtDCa#x^fm=l%
zMBjItGLBb6d_6l%y8N(BDE6}>cYkW)qjc_$12(pA9l`uXrS&8S<sUGSuB@jSz#!|%
z@JptQ@XZcVhTf61L5!nz!>_@8*BZcf05>W^XYCDjADE$<>!PkJG^N?Xv!2A@iNw=I
zxt6PG-x-zba&zc<$ky2W(7x9A4v3=96tR6?MZ2duzZ1#(sgK)ZYk!<-8K$8?FjCcx
z_Hi;#&r#AT=Hmj;odl({WM3T^J<H*g93)76)*muw;@Sf^Pp;;CPi(H=-hVCfN5s7B
zma1lARKP{z`DeS>J(|-^J%!=^>(?AvAGpd4B|fxY8ih_@-@J^uX~$b&{d{blvLyP%
zTwCcK*?IxWq%@lxS)2$6Vt7o!?|CgBSMR_#n_rl(?H3Jx8t6_k1Bh0*C9C4fI>ZG`
zH;#TVJc7I{bHg^wyVP8SdAal2U_s78T>%-lvt2%)!xG9dVJ-%e`mN|Yl;+}6CTZXm
zy6YbkUmi>}ZI{}CU28)ZWEsttw}#h%vwToq;YCQ){Z5C**|3WF{FQ8_gUJt718X-%
znRq`UJVH5I?=s&6fqMZTeeueZ7(?_W4nts^EVm%$czyxHICJL3+*@EkGt<+Ra&gqW
z)*y1{U~fe4clJv9z<500zUJ;A+TK_O#XV_Jn@wU29O*^%7X&+l$H9mF;=+<p_5{9K
zsr4|wOBQR{$F6D`Oq-|nPgXf~a<wzi^`Y{=!A(@mM?l|~pQmjaj(4<lJNA}pCn}!c
z9ro7sRtgOs3$y2u!ORRhBhR`P`E<2xdf4t|!*08%{IW+`p}-|K)}1-(n@1yJ$qy$|
zr3!J`t4kO2U8av)8q&#^Vbw1@H>cDu#c96(#WAwBwlmupz5DX)-L%}UyAm(c1;SOy
z))Tnl%D)aQm)7#SD``{9?V~z*PZ&Vfs8Db*dv`ta?1k%UBUEDmayp{C5mHRqS@t0G
z8sCt@8b9e`%G*=bzh})AN;`qBME{z4w_$>YVO&!RXaBrG2}mGKHiPmjwJf?xE)=M9
zN#bV#-0{q%Jgpv5@8K1aan5OFBcmf?fKVk2b8&45vJyU*0cg!U%?^2Pp3w78`1Gpb
z17X3Si>E2saD>5Qgcl3~0YMQgHef%vm><N$7bJGN@%DowLi{{K5g>0d&j>LaFi6bQ
z0|B;&g#^H5#EhX{a$-n71o%P_C<NjMhx@^xVm=`-h#2C3fpEBo7dR9SLLlI$SUAED
zae9y#%ty=y<OPPp0%5)pa?5ARPRqb7H~w=6Zw~}W!z%=Iy09n$6;xyu6lE2aZ4}fr
zfOj+$)n$Py8bIJeYir^EBS2u@em)VW1R(G~1mGz_Nkc_VLq+-j2n1rZ$teNw9}fsW
zA0G{mz#y;(ODGf`1oHCp@dJ6+Mn##OBDnvH071R|JfJKPxTj4h261W<{;w$r3`4*m
zpwp;$u|WKwvVL$`&p;2a0F73>;{Wp!0z<-OPvau{p9luAC@Esit_7STSd0zK^egYV
G$Ndju&6<1w

literal 0
HcmV?d00001

diff --git a/src/RegularizedProblems.jl b/src/RegularizedProblems.jl
index f50e3eb..ed8ce9f 100644
--- a/src/RegularizedProblems.jl
+++ b/src/RegularizedProblems.jl
@@ -4,6 +4,9 @@ using LinearAlgebra, SparseArrays
 using Random, Requires
 using NLPModels
 using Distributions, Noise
+using FFTW
+using Images
+using Wavelets
 
 include("utils.jl")
 include("types.jl")
@@ -12,6 +15,7 @@ include("lrcomp_model.jl")
 include("matrand_model.jl")
 include("group_lasso_model.jl")
 include("nnmf.jl")
+include("denoising_model.jl")
 
 function __init__()
   @require ProximalOperators = "a725b495-10eb-56fe-b38b-717eba820537" begin
diff --git a/src/denoising_data.jl b/src/denoising_data.jl
new file mode 100644
index 0000000..892557d
--- /dev/null
+++ b/src/denoising_data.jl
@@ -0,0 +1,105 @@
+export generate_HW
+
+# Function to unpad an array
+function unpad(x, n_p, m_p, n)
+  a = n_p - n
+  x = reshape_array(x, (n_p, m_p))
+  unpadded_x = x[(a + 1):end, (a + 1):end]
+  return unpadded_x
+end
+
+# Function to pad an array
+function pad(z, s1, s2)
+  x = cat(zeros(size(z, 1), s1), z, dims = 2)
+  x = cat(x, zeros(size(z, 1), s2), dims = 2)
+  x = cat(x, zeros(s2, size(x, 2)), dims = 1)
+  x = cat(zeros(s1, size(x, 2)), x, dims = 1)
+  return x
+end
+
+# Function to pad a specific array to match suitable dimensions
+function pad_x(z, a)
+  t = cat(zeros(size(z, 1), a), z, dims = 2)
+  t = cat(zeros(a, size(z, 2) + a), t, dims = 1)
+  return t
+end
+
+# Function to create a meshgrid for the kernel
+function meshgrid(x_range, y_range)
+  m = length(x_range)
+  n = length(y_range)
+  x = repeat(x_range, inner = (1, n))
+  y = repeat(y_range', outer = (m, 1))
+  return x, y
+end
+
+# Function to generate a Gaussian kernel
+function my_gaussian_kernel(kernel_size, kernel_sigma)
+  x, y = meshgrid((-kernel_size):kernel_size, (-kernel_size):kernel_size)
+  normal = 1 / (2.0 * pi * kernel_sigma^2)
+  kernel = exp.(-((x .^ 2 + y .^ 2) / (2.0 * kernel_sigma^2))) * normal
+  kernel = kernel / sum(kernel)
+  return kernel
+end
+
+# Function to generate a uniform kernel
+function uniform_kernel(kernel_size)
+  kernel = ones(2 * kernel_size + 1, 2 * kernel_size + 1)
+  kernel = kernel / sum(kernel)
+  return kernel
+end
+
+# Main function to generate H, H_T, W, and W_T functions
+function generate_HW(shape, shape_p, KERNEL_SIZE, KERNEL_TYPE, KERNEL_SIGMA = 1.5)
+  (n, m) = shape
+  (n_p, m_p) = shape_p
+  a = n_p - n
+
+  if KERNEL_TYPE == "gaussian"
+    kernel_h = my_gaussian_kernel(KERNEL_SIZE, KERNEL_SIGMA)
+  elseif KERNEL_TYPE == "uniform"
+    kernel_h = uniform_kernel(KERNEL_SIZE)
+  else
+    println("This KERNEL_TYPE is not defined")
+  end
+
+  sz = (n_p - (2 * KERNEL_SIZE + 1), m_p - (2 * KERNEL_SIZE + 1))
+  kernel_h = pad(kernel_h, div(sz[1], 2), div(sz[1], 2) + 1)
+  kernel_h = ifftshift(kernel_h)
+  fft_h = fft(kernel_h)
+
+  # Function H: Applies a linear transformation H which models the blur to an input x
+  function H(x)
+    x = reshape_array(x, (n, m))
+    x = pad_x(x, a)
+    fft_x = fft(x)
+    x_new = real(ifft(fft_x .* fft_h))
+    return unpad(x_new, n_p, m_p, n)[:]
+  end
+
+  # Function H_T: Applies the transpose of the linear transformation H to an input x
+  function H_T(x)
+    x = reshape_array(x, (n, m))
+    x = pad_x(x, a)
+    fft_x = fft(x)
+    x_new = real(ifft(fft_x .* conj.(fft_h)))
+    return unpad(x_new, n_p, m_p, n)[:]
+  end
+
+  wt = wavelet(WT.haar)
+
+  # ----- Discrete Wavelet Transform (DWT) -----
+
+  # Function W: Applies the DWT to an input x
+  function W(x)
+    return dwt(reshape_array(x, (n, m)), wt, 4)[:]
+  end
+
+  # Function W_T: Applies the inverse DWT to an input x
+  function W_T(x)
+    return idwt(reshape_array(x, (n, m)), wt, 4)[:]
+  end
+
+  # Return the generated functions
+  return H, H_T, W, W_T
+end
diff --git a/src/denoising_model.jl b/src/denoising_model.jl
new file mode 100644
index 0000000..36afefe
--- /dev/null
+++ b/src/denoising_model.jl
@@ -0,0 +1,34 @@
+export denoising_model
+
+include("denoising_data.jl")
+
+function denoing_model(shape, shape_p, KERNEL_SIZE, KERNEL_TYPE, KERNEL_SIGMA = 1.5)
+  sigma = 10^-3
+  cameraman_image = load("images/cameraman.png")
+  x_t = vec(Float64.(cameraman_image))
+  H, H_T, W, W_T = generate_HW(shape, shape_p, KERNEL_SIZE, KERNEL_TYPE, KERNEL_SIGMA)
+  (n, m) = shape
+  b = H(x_t) + randn(n * m) * sigma
+  y = similar(b)
+  z = similar(b)
+
+  function obj(x)
+    y .= W_T(x)
+    y .= H(y)
+    z .= log.((y - b) .^ 2 .+ 1)
+    return sum(z)
+  end
+
+  function grad!(g, x)
+    y = H(W_T(x))
+    term = 1.0 ./ ((y - b) .^ 2 .+ 1)
+    z = term .* (y - b)
+    z .*= 2
+    g .= W(H_T(z))
+    return g
+  end
+
+  x0 = W(b)
+
+  FirstOrderModel(obj, grad!, x0, name = "denoing_model"), x_t
+end
diff --git a/test/runtests.jl b/test/runtests.jl
index 2820353..9a326c6 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -1,5 +1,6 @@
 using LinearAlgebra, Test
 using ADNLPModels, DifferentialEquations, NLPModels, MLDatasets, QuadraticModels
+using Images, FFTW, Wavelets
 using RegularizedProblems
 
 function test_well_defined(model, nls_model, sol)
@@ -166,3 +167,13 @@ end
 end
 
 include("rmodel_tests.jl")
+@testset "denoing_model" begin
+  n, m = 256, 256
+  n_p, m_p = 260, 260
+  kz = 9
+  kt = "gaussian"
+  model, sol = denoing_model((n, m), (n_p, m_p), kz, kt)
+  @test typeof(model) <: FirstOrderModel
+  @test typeof(sol) == typeof(model.meta.x0)
+  @test model.meta.nvar == n * m
+end

From 25e9997e6b6fb2fb08cb34314d5e5b3d22bd41f0 Mon Sep 17 00:00:00 2001
From: MohamedLaghdafHABIBOULLAH <mohamed-laghdaf-2.habiboullah@polymtl.ca>
Date: Thu, 12 Sep 2024 14:37:40 -0400
Subject: [PATCH 02/16] Add denoising problem

---
 src/denoising_data.jl  | 104 ++++++++++++++++++++++++++++++++---------
 src/denoising_model.jl |  19 ++++----
 2 files changed, 93 insertions(+), 30 deletions(-)

diff --git a/src/denoising_data.jl b/src/denoising_data.jl
index 892557d..e0d81e2 100644
--- a/src/denoising_data.jl
+++ b/src/denoising_data.jl
@@ -1,4 +1,21 @@
-export generate_HW
+export generate_uniform_blur, generate_gaussian_blur
+
+"""
+Implementation of the denoising problem described in 
+
+Stella, L., Themelis, A. & Patrinos, P. 
+Forward–backward quasi-Newton methods for nonsmooth optimization problems. 
+Comput Optim Appl 67, 443–487 (2017). https://doi.org/10.1007/s10589-017-9912-y
+
+and adapted from the original implementation in Python by the authors of the paper:
+
+Chouzenoux, E., Martin, S. & Pesquet, JC. 
+A Local MM Subspace Method for Solving Constrained Variational Problems in Image Recovery. 
+J Math Imaging Vis 65, 253–276 (2023). https://doi.org/10.1007/s10851-022-01112-z
+"""
+# non-allocating reshape
+# see https://github.com/JuliaLang/julia/issues/36313
+reshape_array(a, dims) = invoke(Base._reshape, Tuple{AbstractArray, typeof(dims)}, a, dims)
 
 # Function to unpad an array
 function unpad(x, n_p, m_p, n)
@@ -24,7 +41,7 @@ function pad_x(z, a)
   return t
 end
 
-# Function to create a meshgrid for the kernel
+# Function to create a meshgrid for the gaussian kernel
 function meshgrid(x_range, y_range)
   m = length(x_range)
   n = length(y_range)
@@ -36,9 +53,9 @@ end
 # Function to generate a Gaussian kernel
 function my_gaussian_kernel(kernel_size, kernel_sigma)
   x, y = meshgrid((-kernel_size):kernel_size, (-kernel_size):kernel_size)
-  normal = 1 / (2.0 * pi * kernel_sigma^2)
-  kernel = exp.(-((x .^ 2 + y .^ 2) / (2.0 * kernel_sigma^2))) * normal
-  kernel = kernel / sum(kernel)
+  normal = 1 / (2 * pi * kernel_sigma^2)
+  kernel = exp(-((x ^ 2 + y ^ 2) / (2 * kernel_sigma^2))) * normal
+  kernel ./= sum(kernel)
   return kernel
 end
 
@@ -49,31 +66,74 @@ function uniform_kernel(kernel_size)
   return kernel
 end
 
-# Main function to generate H, H_T, W, and W_T functions
-function generate_HW(shape, shape_p, KERNEL_SIZE, KERNEL_TYPE, KERNEL_SIGMA = 1.5)
+# Main function to generate H, H_T, W, and W_T functions for gaussian kernel
+function generate_gaussian_blur(shape, shape_p, KERNEL_SIZE, KERNEL_SIGMA = 1.5)
   (n, m) = shape
   (n_p, m_p) = shape_p
   a = n_p - n
+  
+  kernel_h = my_gaussian_kernel(KERNEL_SIZE, KERNEL_SIGMA)
+
+  sz = (n_p - (2 * KERNEL_SIZE + 1), m_p - (2 * KERNEL_SIZE + 1))
+  kernel_h = pad(kernel_h, div(sz[1], 2), div(sz[1], 2) + 1)
+  kernel_h = FFTW.ifftshift(kernel_h)
+  fft_h = FFTW.fft(kernel_h)
+
+  # Function H: Applies a linear transformation H which models the blur to an input x
+  function H(x)
+    x = reshape_array(x, (n, m))
+    x = pad_x(x, a)
+    fft_x = FFTW.fft(x)
+    x_new = real(FFTW.ifft(fft_x .* fft_h))
+    return unpad(x_new, n_p, m_p, n)[:]
+  end
 
-  if KERNEL_TYPE == "gaussian"
-    kernel_h = my_gaussian_kernel(KERNEL_SIZE, KERNEL_SIGMA)
-  elseif KERNEL_TYPE == "uniform"
-    kernel_h = uniform_kernel(KERNEL_SIZE)
-  else
-    println("This KERNEL_TYPE is not defined")
+  # Function H_T: Applies the transpose of the linear transformation H to an input x
+  function H_T(x)
+    x = reshape_array(x, (n, m))
+    x = pad_x(x, a)
+    fft_x = FFTW.fft(x)
+    x_new = real(FFTW.ifft(fft_x .* conj.(fft_h)))
+    return unpad(x_new, n_p, m_p, n)[:]
   end
 
+  wt = Wavelets.wavelet(Wavelets.WT.haar)
+
+  # ----- Discrete Wavelet Transform (DWT) -----
+
+  # Function W: Applies the DWT to an input x
+  function W(x)
+    return Wavelets.dwt(reshape_array(x, (n, m)), wt, 4)[:]
+  end
+
+  # Function W_T: Applies the inverse DWT to an input x
+  function W_T(x)
+    return Wavelets.idwt(reshape_array(x, (n, m)), wt, 4)[:]
+  end
+
+  # Return the generated functions
+  return H, H_T, W, W_T
+end
+
+# Main function to generate H, H_T, W, and W_T functions for gaussian kernel
+function generate_uniform_blur(shape, shape_p, KERNEL_SIZE)
+  (n, m) = shape
+  (n_p, m_p) = shape_p
+  a = n_p - n
+
+  kernel_h = uniform_kernel(KERNEL_SIZE)
+  
   sz = (n_p - (2 * KERNEL_SIZE + 1), m_p - (2 * KERNEL_SIZE + 1))
   kernel_h = pad(kernel_h, div(sz[1], 2), div(sz[1], 2) + 1)
-  kernel_h = ifftshift(kernel_h)
-  fft_h = fft(kernel_h)
+  kernel_h = FFTW.ifftshift(kernel_h)
+  fft_h = FFTW.fft(kernel_h)
 
   # Function H: Applies a linear transformation H which models the blur to an input x
   function H(x)
     x = reshape_array(x, (n, m))
     x = pad_x(x, a)
-    fft_x = fft(x)
-    x_new = real(ifft(fft_x .* fft_h))
+    fft_x = FFTW.fft(x)
+    x_new = real(FFTW.ifft(fft_x .* fft_h))
     return unpad(x_new, n_p, m_p, n)[:]
   end
 
@@ -81,23 +141,23 @@ function generate_HW(shape, shape_p, KERNEL_SIZE, KERNEL_TYPE, KERNEL_SIGMA = 1.
   function H_T(x)
     x = reshape_array(x, (n, m))
     x = pad_x(x, a)
-    fft_x = fft(x)
-    x_new = real(ifft(fft_x .* conj.(fft_h)))
+    fft_x = FFTW.fft(x)
+    x_new = real(FFTW.ifft(fft_x .* conj.(fft_h)))
     return unpad(x_new, n_p, m_p, n)[:]
   end
 
-  wt = wavelet(WT.haar)
+  wt = Wavelets.wavelet(Wavelets.WT.haar)
 
   # ----- Discrete Wavelet Transform (DWT) -----
 
   # Function W: Applies the DWT to an input x
   function W(x)
-    return dwt(reshape_array(x, (n, m)), wt, 4)[:]
+    return Wavelets.dwt(reshape_array(x, (n, m)), wt, 4)[:]
   end
 
   # Function W_T: Applies the inverse DWT to an input x
   function W_T(x)
-    return idwt(reshape_array(x, (n, m)), wt, 4)[:]
+    return Wavelets.idwt(reshape_array(x, (n, m)), wt, 4)[:]
   end
 
   # Return the generated functions
diff --git a/src/denoising_model.jl b/src/denoising_model.jl
index 36afefe..7431b7e 100644
--- a/src/denoising_model.jl
+++ b/src/denoising_model.jl
@@ -1,12 +1,16 @@
+using Images
+using FFTW
+using Wavelets
 export denoising_model
 
 include("denoising_data.jl")
 
-function denoing_model(shape, shape_p, KERNEL_SIZE, KERNEL_TYPE, KERNEL_SIGMA = 1.5)
+function denoising_model(shape, shape_p, KERNEL_SIZE, KERNEL_SIGMA = 1.5)
   sigma = 10^-3
-  cameraman_image = load("images/cameraman.png")
+  data_path = joinpath(@__DIR__, "..", "images/cameraman.png")
+  cameraman_image = Images.load(data_path)
   x_t = vec(Float64.(cameraman_image))
-  H, H_T, W, W_T = generate_HW(shape, shape_p, KERNEL_SIZE, KERNEL_TYPE, KERNEL_SIGMA)
+  H, H_T, W, W_T = generate_gaussian_blur(shape, shape_p, KERNEL_SIZE, KERNEL_SIGMA)
   (n, m) = shape
   b = H(x_t) + randn(n * m) * sigma
   y = similar(b)
@@ -15,15 +19,14 @@ function denoing_model(shape, shape_p, KERNEL_SIZE, KERNEL_TYPE, KERNEL_SIGMA =
   function obj(x)
     y .= W_T(x)
     y .= H(y)
-    z .= log.((y - b) .^ 2 .+ 1)
+    z .= log.((y .- b) .^ 2 .+ 1)
     return sum(z)
   end
 
   function grad!(g, x)
-    y = H(W_T(x))
-    term = 1.0 ./ ((y - b) .^ 2 .+ 1)
-    z = term .* (y - b)
-    z .*= 2
+    y .= H(W_T(x))
+    z .= 1 ./ ((y .- b) .^ 2 .+ 1)
+    @. z = 2 * z * (y - b)
     g .= W(H_T(z))
     return g
   end

From c2e4e7e28929df07e5cc1a42285d731af810e1a6 Mon Sep 17 00:00:00 2001
From: MohamedLaghdafHABIBOULLAH <laghdaf.hacen@gmail.com>
Date: Tue, 4 Jul 2023 17:55:56 -0400
Subject: [PATCH 03/16] Update Project.toml

Add missing packages
---
 Project.toml | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/Project.toml b/Project.toml
index 58fcc23..1d07a3e 100644
--- a/Project.toml
+++ b/Project.toml
@@ -5,12 +5,15 @@ version = "0.1.1"
 
 [deps]
 Distributions = "31c24e10-a181-5473-b8eb-7969acd0382f"
+FFTW = "7a1cc6ca-52ef-59f5-83cd-3a7055c09341"
+Images = "916415d5-f1e6-5110-898d-aaa5f9f070e0"
 LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
 NLPModels = "a4795742-8479-5a88-8948-cc11e1c8c1a6"
 Noise = "81d43f40-5267-43b7-ae1c-8b967f377efa"
 Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
 Requires = "ae029012-a4dd-5104-9daa-d747884805df"
 SparseArrays = "2f01184e-e22b-5df5-ae63-d93ebab69eaf"
+Wavelets = "29a6e085-ba6d-5f35-a997-948ac2efa89a"
 
 [compat]
 ADNLPModels = "^0.3, 0.4, 0.7"

From fbdf953c9f5ccedc63936bf1e86778187483c48b Mon Sep 17 00:00:00 2001
From: MohamedLaghdafHABIBOULLAH <laghdaf.hacen@gmail.com>
Date: Tue, 4 Jul 2023 19:00:12 -0400
Subject: [PATCH 04/16] Name correctly the function

---
 src/denoising_model.jl | 2 +-
 test/runtests.jl       | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/denoising_model.jl b/src/denoising_model.jl
index 7431b7e..46c4523 100644
--- a/src/denoising_model.jl
+++ b/src/denoising_model.jl
@@ -5,7 +5,7 @@ export denoising_model
 
 include("denoising_data.jl")
 
-function denoising_model(shape, shape_p, KERNEL_SIZE, KERNEL_SIGMA = 1.5)
+function denoising_model(shape, shape_p, KERNEL_SIZE, KERNEL_TYPE, KERNEL_SIGMA = 1.5)
   sigma = 10^-3
   data_path = joinpath(@__DIR__, "..", "images/cameraman.png")
   cameraman_image = Images.load(data_path)
diff --git a/test/runtests.jl b/test/runtests.jl
index 9a326c6..83de8fb 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -172,7 +172,7 @@ include("rmodel_tests.jl")
   n_p, m_p = 260, 260
   kz = 9
   kt = "gaussian"
-  model, sol = denoing_model((n, m), (n_p, m_p), kz, kt)
+  model, sol = denoising_model((n, m), (n_p, m_p), kz, kt)
   @test typeof(model) <: FirstOrderModel
   @test typeof(sol) == typeof(model.meta.x0)
   @test model.meta.nvar == n * m

From 90a4e8afd9d42f5fd3616a05d5e31d72239b682b Mon Sep 17 00:00:00 2001
From: MohamedLaghdafHABIBOULLAH <laghdaf.hacen@gmail.com>
Date: Wed, 9 Aug 2023 17:58:14 -0400
Subject: [PATCH 05/16] Update project.toml and reduce allocations

---
 Project.toml               |  5 +++--
 src/RegularizedProblems.jl | 11 +++++++----
 src/denoising_data.jl      | 14 ++++++++++++++
 src/denoising_model.jl     |  6 +++++-
 test/runtests.jl           |  6 +++---
 5 files changed, 32 insertions(+), 10 deletions(-)

diff --git a/Project.toml b/Project.toml
index 1d07a3e..990d792 100644
--- a/Project.toml
+++ b/Project.toml
@@ -5,8 +5,6 @@ version = "0.1.1"
 
 [deps]
 Distributions = "31c24e10-a181-5473-b8eb-7969acd0382f"
-FFTW = "7a1cc6ca-52ef-59f5-83cd-3a7055c09341"
-Images = "916415d5-f1e6-5110-898d-aaa5f9f070e0"
 LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
 NLPModels = "a4795742-8479-5a88-8948-cc11e1c8c1a6"
 Noise = "81d43f40-5267-43b7-ae1c-8b967f377efa"
@@ -32,6 +30,9 @@ ProximalOperators = "a725b495-10eb-56fe-b38b-717eba820537"
 QuadraticModels = "f468eda6-eac5-11e8-05a5-ff9e497bcd19"
 ShiftedProximalOperators = "d4fd37fa-580c-4e43-9b30-361c21aae263"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
+FFTW = "7a1cc6ca-52ef-59f5-83cd-3a7055c09341"
+Images = "916415d5-f1e6-5110-898d-aaa5f9f070e0"
+Wavelets = "29a6e085-ba6d-5f35-a997-948ac2efa89a"
 
 [targets]
 test = ["ADNLPModels", "DifferentialEquations", "MLDatasets", "ProximalOperators", "QuadraticModels", "ShiftedProximalOperators", "Test"]
diff --git a/src/RegularizedProblems.jl b/src/RegularizedProblems.jl
index ed8ce9f..0b8c34c 100644
--- a/src/RegularizedProblems.jl
+++ b/src/RegularizedProblems.jl
@@ -4,9 +4,6 @@ using LinearAlgebra, SparseArrays
 using Random, Requires
 using NLPModels
 using Distributions, Noise
-using FFTW
-using Images
-using Wavelets
 
 include("utils.jl")
 include("types.jl")
@@ -15,7 +12,6 @@ include("lrcomp_model.jl")
 include("matrand_model.jl")
 include("group_lasso_model.jl")
 include("nnmf.jl")
-include("denoising_model.jl")
 
 function __init__()
   @require ProximalOperators = "a725b495-10eb-56fe-b38b-717eba820537" begin
@@ -46,6 +42,13 @@ function __init__()
   @require QuadraticModels = "f468eda6-eac5-11e8-05a5-ff9e497bcd19" begin
     include("qp_rand_model.jl")
   end
+  @require FFTW = "7a1cc6ca-52ef-59f5-83cd-3a7055c09341" begin
+    @require Wavelets = "29a6e085-ba6d-5f35-a997-948ac2efa89a" begin
+      @require Images = "916415d5-f1e6-5110-898d-aaa5f9f070e0" begin
+        include("denoising_model.jl")
+      end
+    end
+  end
 end
 
 end
diff --git a/src/denoising_data.jl b/src/denoising_data.jl
index e0d81e2..ca8a2ba 100644
--- a/src/denoising_data.jl
+++ b/src/denoising_data.jl
@@ -3,6 +3,7 @@ export generate_uniform_blur, generate_gaussian_blur
 """
 Implementation of the denoising problem described in 
 
+<<<<<<< HEAD
 Stella, L., Themelis, A. & Patrinos, P. 
 Forward–backward quasi-Newton methods for nonsmooth optimization problems. 
 Comput Optim Appl 67, 443–487 (2017). https://doi.org/10.1007/s10589-017-9912-y
@@ -16,6 +17,13 @@ J Math Imaging Vis 65, 253–276 (2023). https://doi.org/10.1007/s10851-022-0111
 # non-allocating reshape
 # see https://github.com/JuliaLang/julia/issues/36313
 reshape_array(a, dims) = invoke(Base._reshape, Tuple{AbstractArray, typeof(dims)}, a, dims)
+=======
+Chouzenoux, E., Martin, S. & Pesquet, JC. 
+A Local MM Subspace Method for Solving Constrained Variational Problems in Image Recovery. 
+J Math Imaging Vis 65, 253–276 (2023). https://doi.org/10.1007/s10851-022-01112-z
+
+"""
+>>>>>>> 11df75c (Update project.toml and reduce allocations)
 
 # Function to unpad an array
 function unpad(x, n_p, m_p, n)
@@ -53,9 +61,15 @@ end
 # Function to generate a Gaussian kernel
 function my_gaussian_kernel(kernel_size, kernel_sigma)
   x, y = meshgrid((-kernel_size):kernel_size, (-kernel_size):kernel_size)
+<<<<<<< HEAD
   normal = 1 / (2 * pi * kernel_sigma^2)
   kernel = exp(-((x ^ 2 + y ^ 2) / (2 * kernel_sigma^2))) * normal
   kernel ./= sum(kernel)
+=======
+  normal = 1 / (2.0 * pi * kernel_sigma^2)
+  kernel = exp.(-((x .^ 2 + y .^ 2) / (2.0 * kernel_sigma^2))) * normal
+  kernel .= kernel / sum(kernel)
+>>>>>>> 11df75c (Update project.toml and reduce allocations)
   return kernel
 end
 
diff --git a/src/denoising_model.jl b/src/denoising_model.jl
index 46c4523..60ec3d6 100644
--- a/src/denoising_model.jl
+++ b/src/denoising_model.jl
@@ -5,7 +5,7 @@ export denoising_model
 
 include("denoising_data.jl")
 
-function denoising_model(shape, shape_p, KERNEL_SIZE, KERNEL_TYPE, KERNEL_SIGMA = 1.5)
+function denoising_model(shape, shape_p, KERNEL_SIZE, KERNEL_SIGMA = 1.5)
   sigma = 10^-3
   data_path = joinpath(@__DIR__, "..", "images/cameraman.png")
   cameraman_image = Images.load(data_path)
@@ -25,7 +25,11 @@ function denoising_model(shape, shape_p, KERNEL_SIZE, KERNEL_TYPE, KERNEL_SIGMA
 
   function grad!(g, x)
     y .= H(W_T(x))
+<<<<<<< HEAD
     z .= 1 ./ ((y .- b) .^ 2 .+ 1)
+=======
+    z .= 1.0 ./ ((y .- b) .^ 2 .+ 1)
+>>>>>>> 11df75c (Update project.toml and reduce allocations)
     @. z = 2 * z * (y - b)
     g .= W(H_T(z))
     return g
diff --git a/test/runtests.jl b/test/runtests.jl
index 83de8fb..c6cc321 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -168,11 +168,11 @@ end
 
 include("rmodel_tests.jl")
 @testset "denoing_model" begin
+@testset "denoising_model" begin
   n, m = 256, 256
   n_p, m_p = 260, 260
-  kz = 9
-  kt = "gaussian"
-  model, sol = denoising_model((n, m), (n_p, m_p), kz, kt)
+  kernel_size = 9
+  model, sol = denoising_model((n, m), (n_p, m_p), kernel_size)
   @test typeof(model) <: FirstOrderModel
   @test typeof(sol) == typeof(model.meta.x0)
   @test model.meta.nvar == n * m

From 048f892f627c6c6fd10f4916a43a49186ff6b9bd Mon Sep 17 00:00:00 2001
From: MohamedLaghdafHABIBOULLAH <laghdaf.hacen@gmail.com>
Date: Fri, 11 Aug 2023 11:20:58 -0400
Subject: [PATCH 06/16] denoising data remove extra allocation

---
 src/denoising_data.jl | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/src/denoising_data.jl b/src/denoising_data.jl
index ca8a2ba..ebe2713 100644
--- a/src/denoising_data.jl
+++ b/src/denoising_data.jl
@@ -67,9 +67,14 @@ function my_gaussian_kernel(kernel_size, kernel_sigma)
   kernel ./= sum(kernel)
 =======
   normal = 1 / (2.0 * pi * kernel_sigma^2)
+<<<<<<< HEAD
   kernel = exp.(-((x .^ 2 + y .^ 2) / (2.0 * kernel_sigma^2))) * normal
   kernel .= kernel / sum(kernel)
 >>>>>>> 11df75c (Update project.toml and reduce allocations)
+=======
+  kernel = exp.(-((x .^ 2 .+ y .^ 2) / (2.0 * kernel_sigma^2))) * normal
+  kernel ./= sum(kernel)
+>>>>>>> a8c1f29 (denoising data remove extra allocation)
   return kernel
 end
 

From 27a1755338c3c6432fc6e2f8abf0db88902d7c12 Mon Sep 17 00:00:00 2001
From: MohamedLaghdafHABIBOULLAH <laghdaf.hacen@gmail.com>
Date: Fri, 11 Aug 2023 11:29:48 -0400
Subject: [PATCH 07/16] remove extra allocation in kernel function

---
 src/denoising_data.jl | 22 ----------------------
 1 file changed, 22 deletions(-)

diff --git a/src/denoising_data.jl b/src/denoising_data.jl
index ebe2713..0bbe2f0 100644
--- a/src/denoising_data.jl
+++ b/src/denoising_data.jl
@@ -3,7 +3,6 @@ export generate_uniform_blur, generate_gaussian_blur
 """
 Implementation of the denoising problem described in 
 
-<<<<<<< HEAD
 Stella, L., Themelis, A. & Patrinos, P. 
 Forward–backward quasi-Newton methods for nonsmooth optimization problems. 
 Comput Optim Appl 67, 443–487 (2017). https://doi.org/10.1007/s10589-017-9912-y
@@ -14,16 +13,6 @@ Chouzenoux, E., Martin, S. & Pesquet, JC.
 A Local MM Subspace Method for Solving Constrained Variational Problems in Image Recovery. 
 J Math Imaging Vis 65, 253–276 (2023). https://doi.org/10.1007/s10851-022-01112-z
 """
-# non-allocating reshape
-# see https://github.com/JuliaLang/julia/issues/36313
-reshape_array(a, dims) = invoke(Base._reshape, Tuple{AbstractArray, typeof(dims)}, a, dims)
-=======
-Chouzenoux, E., Martin, S. & Pesquet, JC. 
-A Local MM Subspace Method for Solving Constrained Variational Problems in Image Recovery. 
-J Math Imaging Vis 65, 253–276 (2023). https://doi.org/10.1007/s10851-022-01112-z
-
-"""
->>>>>>> 11df75c (Update project.toml and reduce allocations)
 
 # Function to unpad an array
 function unpad(x, n_p, m_p, n)
@@ -61,20 +50,9 @@ end
 # Function to generate a Gaussian kernel
 function my_gaussian_kernel(kernel_size, kernel_sigma)
   x, y = meshgrid((-kernel_size):kernel_size, (-kernel_size):kernel_size)
-<<<<<<< HEAD
-  normal = 1 / (2 * pi * kernel_sigma^2)
-  kernel = exp(-((x ^ 2 + y ^ 2) / (2 * kernel_sigma^2))) * normal
-  kernel ./= sum(kernel)
-=======
   normal = 1 / (2.0 * pi * kernel_sigma^2)
-<<<<<<< HEAD
-  kernel = exp.(-((x .^ 2 + y .^ 2) / (2.0 * kernel_sigma^2))) * normal
-  kernel .= kernel / sum(kernel)
->>>>>>> 11df75c (Update project.toml and reduce allocations)
-=======
   kernel = exp.(-((x .^ 2 .+ y .^ 2) / (2.0 * kernel_sigma^2))) * normal
   kernel ./= sum(kernel)
->>>>>>> a8c1f29 (denoising data remove extra allocation)
   return kernel
 end
 

From 44635b9d1fd3d84efce3656efb5700a2e0719b0d Mon Sep 17 00:00:00 2001
From: MohamedLaghdafHABIBOULLAH <laghdaf.hacen@gmail.com>
Date: Fri, 11 Aug 2023 11:43:15 -0400
Subject: [PATCH 08/16] add test for obj and grad

---
 test/runtests.jl | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/test/runtests.jl b/test/runtests.jl
index c6cc321..d9522d0 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -176,4 +176,7 @@ include("rmodel_tests.jl")
   @test typeof(model) <: FirstOrderModel
   @test typeof(sol) == typeof(model.meta.x0)
   @test model.meta.nvar == n * m
+  x = model.meta.x0
+  @test typeof(obj(model, x)) <: Float64
+  @test typeof(grad(model, x)) <: Vector{Float64}
 end

From f7896845bab88d762d1f7a39358da647ed820dd8 Mon Sep 17 00:00:00 2001
From: Dominique <dominique.orban@gmail.com>
Date: Tue, 20 Feb 2024 12:20:54 -0500
Subject: [PATCH 09/16] Update src/denoising_model.jl

---
 src/denoising_model.jl | 4 ----
 1 file changed, 4 deletions(-)

diff --git a/src/denoising_model.jl b/src/denoising_model.jl
index 60ec3d6..7431b7e 100644
--- a/src/denoising_model.jl
+++ b/src/denoising_model.jl
@@ -25,11 +25,7 @@ function denoising_model(shape, shape_p, KERNEL_SIZE, KERNEL_SIGMA = 1.5)
 
   function grad!(g, x)
     y .= H(W_T(x))
-<<<<<<< HEAD
     z .= 1 ./ ((y .- b) .^ 2 .+ 1)
-=======
-    z .= 1.0 ./ ((y .- b) .^ 2 .+ 1)
->>>>>>> 11df75c (Update project.toml and reduce allocations)
     @. z = 2 * z * (y - b)
     g .= W(H_T(z))
     return g

From 5ce08057b94c7b4541a2757a87093b1a6d214767 Mon Sep 17 00:00:00 2001
From: MohamedLaghdafHABIBOULLAH <mohamed-laghdaf-2.habiboullah@polymtl.ca>
Date: Thu, 12 Sep 2024 15:00:07 -0400
Subject: [PATCH 10/16] Update runtests.jl

---
 test/runtests.jl | 1 -
 1 file changed, 1 deletion(-)

diff --git a/test/runtests.jl b/test/runtests.jl
index d9522d0..d647139 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -167,7 +167,6 @@ end
 end
 
 include("rmodel_tests.jl")
-@testset "denoing_model" begin
 @testset "denoising_model" begin
   n, m = 256, 256
   n_p, m_p = 260, 260

From 2f62ffddc2dac68c39a7669a06e3c2b3621f478f Mon Sep 17 00:00:00 2001
From: MohamedLaghdafHABIBOULLAH <mohamed-laghdaf-2.habiboullah@polymtl.ca>
Date: Thu, 12 Sep 2024 16:15:13 -0400
Subject: [PATCH 11/16] resoudre projedct toml

---
 Project.toml               | 12 +++++++++---
 src/RegularizedProblems.jl |  9 ++-------
 2 files changed, 11 insertions(+), 10 deletions(-)

diff --git a/Project.toml b/Project.toml
index 990d792..5003567 100644
--- a/Project.toml
+++ b/Project.toml
@@ -4,10 +4,16 @@ authors = ["Dominique Orban <dominique.orban@gmail.com>, Robert Baraldi <robertj
 version = "0.1.1"
 
 [deps]
+ADNLPModels = "54578032-b7ea-4c30-94aa-7cbd1cce6c9a"
+DifferentialEquations = "0c46a032-eb83-5123-abaf-570d42b7fbaa"
 Distributions = "31c24e10-a181-5473-b8eb-7969acd0382f"
+FFTW = "7a1cc6ca-52ef-59f5-83cd-3a7055c09341"
+Images = "916415d5-f1e6-5110-898d-aaa5f9f070e0"
 LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
+MLDatasets = "eb30cadb-4394-5ae3-aed4-317e484a6458"
 NLPModels = "a4795742-8479-5a88-8948-cc11e1c8c1a6"
 Noise = "81d43f40-5267-43b7-ae1c-8b967f377efa"
+QuadraticModels = "f468eda6-eac5-11e8-05a5-ff9e497bcd19"
 Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
 Requires = "ae029012-a4dd-5104-9daa-d747884805df"
 SparseArrays = "2f01184e-e22b-5df5-ae63-d93ebab69eaf"
@@ -25,14 +31,14 @@ julia = "^1.6.0"
 [extras]
 ADNLPModels = "54578032-b7ea-4c30-94aa-7cbd1cce6c9a"
 DifferentialEquations = "0c46a032-eb83-5123-abaf-570d42b7fbaa"
+FFTW = "7a1cc6ca-52ef-59f5-83cd-3a7055c09341"
+Images = "916415d5-f1e6-5110-898d-aaa5f9f070e0"
 MLDatasets = "eb30cadb-4394-5ae3-aed4-317e484a6458"
 ProximalOperators = "a725b495-10eb-56fe-b38b-717eba820537"
 QuadraticModels = "f468eda6-eac5-11e8-05a5-ff9e497bcd19"
 ShiftedProximalOperators = "d4fd37fa-580c-4e43-9b30-361c21aae263"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
-FFTW = "7a1cc6ca-52ef-59f5-83cd-3a7055c09341"
-Images = "916415d5-f1e6-5110-898d-aaa5f9f070e0"
 Wavelets = "29a6e085-ba6d-5f35-a997-948ac2efa89a"
 
 [targets]
-test = ["ADNLPModels", "DifferentialEquations", "MLDatasets", "ProximalOperators", "QuadraticModels", "ShiftedProximalOperators", "Test"]
+test = ["ADNLPModels", "DifferentialEquations", "MLDatasets", "ProximalOperators", "QuadraticModels", "ShiftedProximalOperators", "Test", "Images", "FFTW", "Wavelets"]
diff --git a/src/RegularizedProblems.jl b/src/RegularizedProblems.jl
index 0b8c34c..6f1073b 100644
--- a/src/RegularizedProblems.jl
+++ b/src/RegularizedProblems.jl
@@ -4,6 +4,7 @@ using LinearAlgebra, SparseArrays
 using Random, Requires
 using NLPModels
 using Distributions, Noise
+using Images, FFTW, Wavelets
 
 include("utils.jl")
 include("types.jl")
@@ -12,6 +13,7 @@ include("lrcomp_model.jl")
 include("matrand_model.jl")
 include("group_lasso_model.jl")
 include("nnmf.jl")
+include("denoising_model.jl")
 
 function __init__()
   @require ProximalOperators = "a725b495-10eb-56fe-b38b-717eba820537" begin
@@ -42,13 +44,6 @@ function __init__()
   @require QuadraticModels = "f468eda6-eac5-11e8-05a5-ff9e497bcd19" begin
     include("qp_rand_model.jl")
   end
-  @require FFTW = "7a1cc6ca-52ef-59f5-83cd-3a7055c09341" begin
-    @require Wavelets = "29a6e085-ba6d-5f35-a997-948ac2efa89a" begin
-      @require Images = "916415d5-f1e6-5110-898d-aaa5f9f070e0" begin
-        include("denoising_model.jl")
-      end
-    end
-  end
 end
 
 end

From 377039360f00950bd801745ab935332d1864fc06 Mon Sep 17 00:00:00 2001
From: MohamedLaghdafHABIBOULLAH
 <81633807+MohamedLaghdafHABIBOULLAH@users.noreply.github.com>
Date: Mon, 4 Nov 2024 13:46:55 -0500
Subject: [PATCH 12/16] Update src/denoising_data.jl

Co-authored-by: Dominique <dominique.orban@gmail.com>
---
 src/denoising_data.jl | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/denoising_data.jl b/src/denoising_data.jl
index 0bbe2f0..59dce60 100644
--- a/src/denoising_data.jl
+++ b/src/denoising_data.jl
@@ -16,7 +16,7 @@ J Math Imaging Vis 65, 253–276 (2023). https://doi.org/10.1007/s10851-022-0111
 
 # Function to unpad an array
 function unpad(x, n_p, m_p, n)
-  a = n_p - n
+  a = n_p - n + 1
   x = reshape_array(x, (n_p, m_p))
   unpadded_x = x[(a + 1):end, (a + 1):end]
   return unpadded_x

From 5b55407e19336b680ff37246963507e9e682d5d5 Mon Sep 17 00:00:00 2001
From: MohamedLaghdafHABIBOULLAH
 <81633807+MohamedLaghdafHABIBOULLAH@users.noreply.github.com>
Date: Mon, 4 Nov 2024 13:47:05 -0500
Subject: [PATCH 13/16] Update src/denoising_data.jl

Co-authored-by: Dominique <dominique.orban@gmail.com>
---
 src/denoising_data.jl | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/denoising_data.jl b/src/denoising_data.jl
index 59dce60..b7236d3 100644
--- a/src/denoising_data.jl
+++ b/src/denoising_data.jl
@@ -18,7 +18,7 @@ J Math Imaging Vis 65, 253–276 (2023). https://doi.org/10.1007/s10851-022-0111
 function unpad(x, n_p, m_p, n)
   a = n_p - n + 1
   x = reshape_array(x, (n_p, m_p))
-  unpadded_x = x[(a + 1):end, (a + 1):end]
+  unpadded_x = x[a:end, a:end]
   return unpadded_x
 end
 

From 4937015ae26605b5c43bc5b0762c4bcf6955f019 Mon Sep 17 00:00:00 2001
From: MohamedLaghdafHABIBOULLAH
 <81633807+MohamedLaghdafHABIBOULLAH@users.noreply.github.com>
Date: Mon, 4 Nov 2024 13:47:30 -0500
Subject: [PATCH 14/16] Update src/denoising_data.jl

Co-authored-by: Dominique <dominique.orban@gmail.com>
---
 src/denoising_data.jl | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/denoising_data.jl b/src/denoising_data.jl
index b7236d3..02ec3f9 100644
--- a/src/denoising_data.jl
+++ b/src/denoising_data.jl
@@ -50,7 +50,7 @@ end
 # Function to generate a Gaussian kernel
 function my_gaussian_kernel(kernel_size, kernel_sigma)
   x, y = meshgrid((-kernel_size):kernel_size, (-kernel_size):kernel_size)
-  normal = 1 / (2.0 * pi * kernel_sigma^2)
+  normal = 1 / (2 * pi * kernel_sigma^2)
   kernel = exp.(-((x .^ 2 .+ y .^ 2) / (2.0 * kernel_sigma^2))) * normal
   kernel ./= sum(kernel)
   return kernel

From 72874a324b0613b12978d4a960cd6149ea4be84d Mon Sep 17 00:00:00 2001
From: MohamedLaghdafHABIBOULLAH
 <81633807+MohamedLaghdafHABIBOULLAH@users.noreply.github.com>
Date: Mon, 4 Nov 2024 13:47:47 -0500
Subject: [PATCH 15/16] Update src/denoising_model.jl

Co-authored-by: Dominique <dominique.orban@gmail.com>
---
 Project.toml           | 7 -------
 src/denoising_model.jl | 3 ---
 2 files changed, 10 deletions(-)

diff --git a/Project.toml b/Project.toml
index 5003567..238102b 100644
--- a/Project.toml
+++ b/Project.toml
@@ -4,20 +4,13 @@ authors = ["Dominique Orban <dominique.orban@gmail.com>, Robert Baraldi <robertj
 version = "0.1.1"
 
 [deps]
-ADNLPModels = "54578032-b7ea-4c30-94aa-7cbd1cce6c9a"
-DifferentialEquations = "0c46a032-eb83-5123-abaf-570d42b7fbaa"
 Distributions = "31c24e10-a181-5473-b8eb-7969acd0382f"
-FFTW = "7a1cc6ca-52ef-59f5-83cd-3a7055c09341"
-Images = "916415d5-f1e6-5110-898d-aaa5f9f070e0"
 LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
-MLDatasets = "eb30cadb-4394-5ae3-aed4-317e484a6458"
 NLPModels = "a4795742-8479-5a88-8948-cc11e1c8c1a6"
 Noise = "81d43f40-5267-43b7-ae1c-8b967f377efa"
-QuadraticModels = "f468eda6-eac5-11e8-05a5-ff9e497bcd19"
 Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
 Requires = "ae029012-a4dd-5104-9daa-d747884805df"
 SparseArrays = "2f01184e-e22b-5df5-ae63-d93ebab69eaf"
-Wavelets = "29a6e085-ba6d-5f35-a997-948ac2efa89a"
 
 [compat]
 ADNLPModels = "^0.3, 0.4, 0.7"
diff --git a/src/denoising_model.jl b/src/denoising_model.jl
index 7431b7e..3f6e379 100644
--- a/src/denoising_model.jl
+++ b/src/denoising_model.jl
@@ -1,6 +1,3 @@
-using Images
-using FFTW
-using Wavelets
 export denoising_model
 
 include("denoising_data.jl")

From 63e1c638f276ddfde6a0fc1022a285389e859395 Mon Sep 17 00:00:00 2001
From: MohamedLaghdafHABIBOULLAH <mohamed-laghdaf-2.habiboullah@polymtl.ca>
Date: Thu, 20 Feb 2025 17:04:01 -0500
Subject: [PATCH 16/16] Correct Project.toml and require to load the model if 
 the required dependencies have been imported

---
 Project.toml               | 2 +-
 src/RegularizedProblems.jl | 9 +++++++--
 test/runtests.jl           | 3 ++-
 3 files changed, 10 insertions(+), 4 deletions(-)

diff --git a/Project.toml b/Project.toml
index 238102b..42644f0 100644
--- a/Project.toml
+++ b/Project.toml
@@ -34,4 +34,4 @@ Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
 Wavelets = "29a6e085-ba6d-5f35-a997-948ac2efa89a"
 
 [targets]
-test = ["ADNLPModels", "DifferentialEquations", "MLDatasets", "ProximalOperators", "QuadraticModels", "ShiftedProximalOperators", "Test", "Images", "FFTW", "Wavelets"]
+test = ["ADNLPModels", "DifferentialEquations", "FFTW", "Images", "MLDatasets", "ProximalOperators", "QuadraticModels", "ShiftedProximalOperators", "Test", "Wavelets"]
diff --git a/src/RegularizedProblems.jl b/src/RegularizedProblems.jl
index 6f1073b..0b8c34c 100644
--- a/src/RegularizedProblems.jl
+++ b/src/RegularizedProblems.jl
@@ -4,7 +4,6 @@ using LinearAlgebra, SparseArrays
 using Random, Requires
 using NLPModels
 using Distributions, Noise
-using Images, FFTW, Wavelets
 
 include("utils.jl")
 include("types.jl")
@@ -13,7 +12,6 @@ include("lrcomp_model.jl")
 include("matrand_model.jl")
 include("group_lasso_model.jl")
 include("nnmf.jl")
-include("denoising_model.jl")
 
 function __init__()
   @require ProximalOperators = "a725b495-10eb-56fe-b38b-717eba820537" begin
@@ -44,6 +42,13 @@ function __init__()
   @require QuadraticModels = "f468eda6-eac5-11e8-05a5-ff9e497bcd19" begin
     include("qp_rand_model.jl")
   end
+  @require FFTW = "7a1cc6ca-52ef-59f5-83cd-3a7055c09341" begin
+    @require Wavelets = "29a6e085-ba6d-5f35-a997-948ac2efa89a" begin
+      @require Images = "916415d5-f1e6-5110-898d-aaa5f9f070e0" begin
+        include("denoising_model.jl")
+      end
+    end
+  end
 end
 
 end
diff --git a/test/runtests.jl b/test/runtests.jl
index d647139..ddefabc 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -166,7 +166,6 @@ end
   @test all(model.meta.x0 .== 0)
 end
 
-include("rmodel_tests.jl")
 @testset "denoising_model" begin
   n, m = 256, 256
   n_p, m_p = 260, 260
@@ -179,3 +178,5 @@ include("rmodel_tests.jl")
   @test typeof(obj(model, x)) <: Float64
   @test typeof(grad(model, x)) <: Vector{Float64}
 end
+
+include("rmodel_tests.jl")
\ No newline at end of file