From 9ef1caa0d5adcc7d6274d4b530037f6cd265e7f3 Mon Sep 17 00:00:00 2001
From: Jody Klymak <jklymak@gmail.com>
Date: Tue, 6 Jul 2021 08:29:11 -0700
Subject: [PATCH 1/2] FIX: fix twoslope norm for colorbars

---
 lib/matplotlib/colors.py                      |  13 +++++++----
 .../test_colorbar/colorbar_twoslope.png       | Bin 5886 -> 5932 bytes
 lib/matplotlib/tests/test_colorbar.py         |   5 +++--
 tutorials/colors/colormapnorms.py             |  21 ++++++++++++++----
 4 files changed, 29 insertions(+), 10 deletions(-)

diff --git a/lib/matplotlib/colors.py b/lib/matplotlib/colors.py
index e0c42c5b69d5..53ac34dfa3db 100644
--- a/lib/matplotlib/colors.py
+++ b/lib/matplotlib/colors.py
@@ -1300,9 +1300,13 @@ def __call__(self, value, clip=None):
 
         if not self.vmin <= self.vcenter <= self.vmax:
             raise ValueError("vmin, vcenter, vmax must increase monotonically")
+        # must linearly extrapolate for ticks on colorbars that go
+        # beyond vmin and vmax:
+        min = self.vmin - (self.vcenter - self.vmin)
+        max = self.vmax + (self.vmax - self.vcenter)
         result = np.ma.masked_array(
-            np.interp(result, [self.vmin, self.vcenter, self.vmax],
-                      [0, 0.5, 1.]), mask=np.ma.getmask(result))
+            np.interp(result, [min, self.vcenter, max],
+                      [-0.5, 0.5, 1.5]), mask=np.ma.getmask(result))
         if is_scalar:
             result = np.atleast_1d(result)[0]
         return result
@@ -1313,8 +1317,9 @@ def inverse(self, value):
         (vmin,), _ = self.process_value(self.vmin)
         (vmax,), _ = self.process_value(self.vmax)
         (vcenter,), _ = self.process_value(self.vcenter)
-
-        result = np.interp(value, [0, 0.5, 1.], [vmin, vcenter, vmax])
+        min = vmin - (vcenter - vmin)
+        max = vmax + (vmax - vcenter)
+        result = np.interp(value, [-0.5, 0.5, 1.5], [min, vcenter, max])
         return result
 
 
diff --git a/lib/matplotlib/tests/baseline_images/test_colorbar/colorbar_twoslope.png b/lib/matplotlib/tests/baseline_images/test_colorbar/colorbar_twoslope.png
index 27d9227dcfe00c31b6ab1cb05484412a646662b1..b92103ae134756ab315f9c36ff1ac4434031da39 100644
GIT binary patch
literal 5932
zcmd^DdpwkB-@j&z9FmA~h*DCk+R0%^ve9moWz??TuyPhn22&;!#%Rkbr=D6-qqN$_
zwlpe2GnmnVM69NVBn?K?I}N7N&^SETeGkJv@B4l}@Bfd_rw`5C_kCU0@Ap0Z?$dib
zb}mvkR0jYoa^1D#M*y&N05CnO3*aX<ZLj*_FAu7-H`Oybj2aho<NzQBQ7PnTDwz~)
z5_{lC3@JLw8c)Dm;VIEER7)$HEgQmZLTzk<ErP>r!#0{6q*5s{_M0|ED(=EZ9|_;2
z`cq0Oe6kv4mro1;i-VAVm_HngNB}H5>AGY4-uQfRcU_cG=g|c{p8wQzvtF?@(k@o_
znEiIG(4{KJvK@B3UgdW4cRxp`*CtN3jd)J=_KMDUk@JIZD-Idx*fjq6<b<jHr?VS9
zE3dD=B1>|r3sf>&arB1=g27h_oQq5nJnx@vct5y>fD1qs>@mP-fvZ+n6X0|J=LG0a
zROtZZXafvbZwjzruM5C|KQjK)E*#kDrZ@6yrcggxP@6f}TCkH(aLZNVogwmdhXvg1
z+)V*4U9{S%re7AX?Ca|*9BbXOb!&Qf>y^t_SGVs18%I8iyIw?9y5(YcI(G!a!meoU
z<i_OB$kpKiNjic@dnUqb;w7tT9~lDG6jwF6@)gJA724X<)X%wXHlloHKnDMr*{g`W
zc3<CzYyl@Nt!Zh8*Hp2}hkI$3Q#Q1Oj~roHVs?7^7(R8ZVeCNu(vlBs_|2PS+HyTT
zJ#65Q*ua$R5CMziwHz?Qo~{b2x$9gat4a5cOL}nJ?dL~Hr_H_#uwmG37^^A2c*%P_
zxzN14<vAAc!mlZ*g2r;<L*~`C7cOtX+I0ZyWW`{fCf5CNOvj_*af_-xgo`~LoNZ=O
zaMfG<%u9ZS=xEf(ST`uX%FwomfneG-u-CNbK`%u#$`Tcd*o7~>RZIgxk4Kl-wxzW-
zn$;auNp;On(=-8V_&a=l84MYwa0NRB(#@r!sW+>Tg^LU^(xl`K;M_0B8ZU77jUo_a
zUQ-%VZ4g>!H3*+}Bv)-!2kkv<C2uS2+O5XCoiD_(m9gJNR(4%?!_rrd*<tbb9uq<0
zNv~4>N1V9fDApv6OLYYxI}A6Ws>#Fkn2-4Injzv%E>=)7A{_Dx!Zjc*jN$(4Rr9t$
zW~>pK930TcC_y%;Xn}KWn9wX_JKjo<jUai>b?yjKq1gMtLfFtDciJ#Gdg#mh+7w^P
zrrkdy7wSf1|1tf{8oK~82|281^}-jbcHbcPT4KYpv@Ia_sL1-Bne@fKJh<7xVWqP~
zXAXrJ<9Iq_vXA)MsE+ZvEqcP^yFVH+N86X(9+KjNyGIv^zW#c3`{Zlq5>Bm62mG|4
zW-(u1(%Y0+T0uIKb!0q#IF?g0`bE@HDtyY=Y<B(j5_lR)(IjnUl0cAnQ<jUCmB3hg
z=`(*T{i)QLS34Z7?DRnk*-R{dhp!e$TC$4<Zwf6VT_Y{j{Bm$#b>6x26b+5pEi)qv
z&F!AvQ#0*(k93^bO=w$&<Lv|PtqL@_^DjhVN&9Gf9N!?9AVqcI%@*oXrXSD29l(;Y
zlmhDvdHTD(7e$2P6dhQrH2@XqtUI63R<0ZYc492HB6TBejzs?3)u?{^y+aU`t+xQe
z<xvCA-G$ew25-m!kBv+NXTPvbvDgXMAgk}>NhVdtD0Z)o-xIOlvRS`d?fdqj3K%Ny
zjuK8xObomf$GYV{JdHjr5*C1>MeW)z;0T7V4a*oEN2leb3FEtX-db?WpmbO)o7xA~
zES_6Q%sgEYsIyB{M(67QSjE1duPXCC$;l7_8k=-xLBNkHq%7)DOApFH>qcT6RByp-
zyq8;4H&?i>ine8BEJ<bZH+gZ#^z%+695X~v@klV~P&?jcFL=@$?9D!zKzvY6Ra^Yi
z#{@lX%l)jloX>=k%8InK-rzT*ek^HtE4OR6wA*e+wI5r(ssf>U;+?)2uyE)N32inr
z61Eo`SXbk)=!yc4v7S}4K)qwWUr8SucpPB^zx))#sRZNRsyIuC-VEQr81aa8lcI0-
zF!)*$l&BE?`Tu?1N%L>8M!{YYnO>TfCXQV`I5@buBdXGOEV}*%EIf#NIV^3HPCo5m
zCm%z4-+aS`2l1*-@0V)={Ns%pO2!fGtcDPy>;^{ZG$l?Klx*kguE2G_U3_3r&GCLz
zCFkz)L!55B-x4svYtW`NI!9d@F|9s6y8d-dLR^ekDnk`Q3tSA5D*>>Y0ltgxUynAx
zR&0lu3B6Qd-i3Q+yRIbFW{KiVcB+?9WiSO+4T>@d&pYKo({!?W8m1!-^Q1c7J3Tgw
zb22sv6%AFoTxytKzry{92*f%C)MvzoUrzJ^WJX$=osnjx|1cZRW^MR0^Qacs-80Dj
za=>A@<pN&l!MDKs-LI){YP#+g|5yCsyRi9&_zqverR1U2t5eg1-Yr={*w2zq`c9Ym
zMJI7N)o0>mqU%lZzlviML{Ja{8W@|SFPNvEUxJ|2GZNb=FS=-^=<X1oq=nnm`eg~S
zTJBQN_+b6+88+F92fnPoSMVI-_AxWEwDg?J^{Ju;e6npDS=}~8_f-?$1lGc~9awz_
z%Se7%VF7+xb>Xt&yaRg+aNqhP+QQdmVaO$zW)H|jR3SqlpsZ1?FmxJ-zdut$at8e(
zf&g316Je=}$1pp|g`J7VRc`hBUs(X9;SQ<mjwMT*f<QGiL`T(nN8+ps5ILurSJS1E
z*MIjUsPvDSq-3t<h$m~neg(O*%g)h}doPyLeLvaK!x>6`6`#h<hSFHuKG}pECe*_Q
zJ0xd5vZBP*q<Q4z4BR^*Gz<`$2(R_znfA??#vREIlx757%Q0usZ~5hs1m)#hzJ~;y
zxiuc^DC^>$nG%J&<sL2~>F45m{e^G-`OM7WU^U{~Z{8r0;HhJrwrIL@`M!y=!A=^x
zG)qr=di+$akEb0rDuZ5zXbyDigX*b?#2EHZ-e#t=8fZ)OeSvwt@`(qse<m~YhRMSm
z*rOgfli(?!6{SMA>Ld_5J^VRPzGlJRSuP&^uOp5(P+9epTjg?wQy%OEh_i<vtsO!@
z>idVGhwDQXI(b_xaAT#x{BxK_qs540Ij5$B^Upxz3ymq%9+6klUtUQt^jtc56YN@n
z6wKO&*ozNT*&dE$4%=9a^>@&<g%bS+p=3bh_{vxrjOeC~W#8%&M~PEkBJm0b96w3Q
z=x+EQb->Ere-?X5hPpY72(Dlg!uh{WXQiy#WZJWn0@i}FDTot~Z45;1gvfU(kHy3|
zBHcTzp<%7C5L<h?qD^x`c!Q*#odUazI!mUu#k}i@F-bYEb{|5)hHKW?tw=}D%TICW
zqrcoiJE9{u_FUB1zh56>1||IM)EBR|$uUUEDn3EeEtkSAZo0l9{x9mW+mOt@lCsO~
zQ)Dmyfc-b6;n9ZevKT$>;;G>$1&Pw$^rTWz3~)>`ZD>fS>=NJX2;kC;qhD8>n`h=g
zkkPk7ZF?)58i?rTIv_L(yTpPlHC_D$WipE3deWKOCYTEl0fCIgvu6HIR@_fJg8J5B
zcJ22=>ozOkW5p*t+I#CC*JfDu_u#UM=&)inY?_se932J_9U+U5kpfkE$=qW}|NdCm
zX2h98-_H+11|0fDR7gE_hp^jx%t~_RGScBo!5a^v8%N~D%JbLE!;iQVs90>FH{V`>
zOfyJ4CH9767??pczlOM5F8jUe9JJ01`m8ItpE3<B;`O87!%h6p@LG?TV+BUG7}&0Z
z{!>IkE7ymevDJoOMOBj?N-c)k`@(LcU)|5`6XcNn(z44eJY;Q<m!6v!rIh3?<)(W}
z%D$0x$GPQ#&1fU?nfeP~#+yk$5Kw2c@DDJPo@ayv0wmNzfWjD!mIJnDJLt!_@Misk
z<+!lwccu9I!QN&_{<*zyeuzSfhrn9{Le6d0(k-9Z9`VNJ70iXw)Wc_fyQ=JcWG0rn
zLmOWk0r+#l)Y`qnt8ub_E!cKn5}zTI#D5&^rFh>m1XYX}xl2MjMWZ`pHJYK3ON+fq
z`?@)Cn>%RS0QMR%mXZy3fj?r6rJY1`|Mez7S22JGgGwszO~#Ch7?{dS+-&Dzxs2TZ
z6_S1oks%b#Um?AQPK$W*)<Pec3*ZJ!^VAb~|3(;A5Zyr1$6Q}Eei(B#Syz1WE@Es5
zkV59dB-+t+lu<r6t++OToxvdrpDle=Y*2auc1_6ksv~TmRowm@z7+=QqmErU(;V-Q
zt*>Q*q}$OpS)LlPn2!J{zJa<@AhLKnIv~#<2{b}{(w6V5*2v*oDAIlit{EK@w2=xb
z%zpb{@*EJf-UuW6s_N6+0gx#NSp{B?a==Td%Q&GGX9k$z3QEluE(2}L>_fxyn1v7W
zjd7O?5HKL)iRrBRLrS3Q8DP%N3*V`vj3A?QNONSp!&-zG6r=h@2XA8fF>|N8XJ~rq
z1pal78T}pV0dTsT+wNOng(4O=zeDq$RTvVOLI0=&)62@Kac<wvR$4XHAnwb+x>i_p
z|K!@fm#bITPQL4K%N@C@w}Cc2^_~$0!^@|lfKpCvz{D8&-IAfvr=4ss&2Od;l|kMe
zcQm?y$3&t0ArLK>QYB1R75P+P&~W3RF$DZLvdv}!(b><zm^aa0lcxme=lRH7#=K2g
zbX6V>)<A`tcMf+UVsiIpP0n(ze>$pYT?Y0x@Sxc_a;uQZ;b`&&+hnA^|Jyhl>iGhb
zF+AKYk5o;wIo6KYru)9%n9Hpkmic1N&PcO*2e4nT_M87}u-(g{H@c~eXu~*ivwKgM
zIQHE+(4$f>k%nVvFYkU5&|CkX!!4=$-xOVF=j?gwd|dfL`6vo~a9%lP+X{t<+4EIy
zo27-k_fQN=Fvu$*WRSHco-%ea?Lo(zFvFt4vI*Ggd2&MZY}4;hwL@%nzG69~OiyJe
zXNSd9Pz`u1-xU|s!U;wtX|gOAx?pMehUDSe@kTG?@7}R)J()k;s4i9Vz3tf%xoQ$L
zEi=Dso9qu$(8)#REE>nHGed%a5i&xq0ni-y77q!!I36-H{cY0`V4X1!9jWT!d^sm9
zvy(jx>2v)@hJO@ncADDC7zzRFF*n*~5KR=cLLz{*hUH!5RZ&`?tW04<&4e5=_R89s
z1;Y#*?tr&Oym^FoKc%w0OF17U?fz`KV^T&0F26wGcnYrnZK4(0?YO4&SP3PpVSO6Y
z1Zc)$(Bx#bO+9PN9C<I+|3=1<7s*FAWL%jUFO``j_LQLaNbmRbhCEglLO;oeWHfX>
zf_^n}wb$-zIANp;M-!xyxQ0Q3AMGnq-+0Z1E5N&FQYO{bww(APSjNJ|XGq=jEJDPi
zKMcT+Vdk6inR;JVRu(VmbDn5bqisUobhWg!c<fmk@!UP2Z~Dv{o2-!PB$+Mp(u>I7
z=+_b+XJ-`5>8?k!VJ?2qRq{H?e%Q}oSWvWeG+{H*vpRWtn}2PhL8NSaUtrSM$SOP@
ze`Y!#S>^NlPPO&HnN+Z_JaV~!CKEq9ka##rAank3x;n0)CBds+CVl5I{yK@(BAAxG
z@3gMV^s!9=iZ`2&SKbt{<0royrTSG;&39)>hxo^apTtjZyDZIWCSx&jsQ;(D9BrP)
XFiMMfoLl5;0RFf-d+cC+8+hVh9%L>L

literal 5886
zcmd^DdpMMN|NhQ2ITUJY<*<sS@<!`aijuK!NJ8&Ar7|U_4m6a>OioR2{i2B8(m|5m
zZH>f^)Y@T)X)zSCn$;wS5m6e9(=<*qe&6R|X4>7~dtJZ(->%EmFwggSK8O3hKlkUk
z_z#aA>Wg$20RYr@x;T0QfTaO|=~7XGf3a?Ee-HoLLw52bdqsqjDThvm0QW=W$nXeq
z_|ZUv=#Z09M<YlZ%r=@?nMFoKk!@_OO%DfIZwlEEWF5RQ*euw}!phRZ+RA3V!4Wb!
zGHQ#t`EmJivxt+2&EN1VUcpDGM!M{c0$|A@<OlPBkb4w>74AD7zw@Tv8SH2x>9w9x
z?(%BcVA818^7A%}0Jq;A);a*?)46u%DOjva-G$_VC+c0bb?1U_`n3I%`RQM^-PsQ_
z7k;aI$<ep2%~P{UVt>!#2-azp@5T2KFH4y$)0}hw?&(NgEd*dVL>1<I84oZ(*9c(2
zMrVKnM^b<a_)QzL6x_4K8UTkNoCBb7RQ}s8tk}n*)rQ}uQzcWU>|#H!XEFA(^nh<c
z`Z9uK+-qjIYDu#=yKr`NsH3c;q~w9rE+;)r`ipwi(kct@Go%~Fz!y0=5M7p254!NJ
zLh;DV<XF>aDgSA#7XM7tGD5ANq}0R1qkFK8sUJ7cTYq!pQ%nA*rp8emdr{NfZOKVV
zXSPmS-i;RAq3;gqUqqr5^>rHt`cqElS0wBSd0OmL!nEdG=eN~s9xO|~)h!;W&8%FR
z-5cibU)^NMsjTMLT)X6RK+Cd5$d_7vo{kS^wq>RXq`lw3$oRTv-ik)tnR`oa@6Mk+
z!*0qIzMf9rVi=fmipT5|P)bL`n0?tW1X7JSdXS(MRuL^U@3`q_HSW&5XsmhpqI<ug
zYjt#)!~k5a-O}a@K-{J+G;qa@?-XvLdpX{*puBb08VI4Qtz=@jiF0G-aB%WK3Jv1y
z{sI;7JltdlThRWx&JyD9N($a0O7WF8X2TAEqXnnXGJ<ew5Qr-sfD1jDrI98*m;mv(
zItbUGhupbH^}OwC6tv32t6nxW^~}u7^oia@`aRMG6F&<y0vc6r4UmEG-N3+lkKMZ?
zQaeaY@=h@PFR@YA?Q`uob}V(I+o}BYSYO{8p?KF#9Egl?W{WM<wyRY=BI|%i<1$(X
zWQBuOe6Kc-YC#ZdLq;H3A^}5ExNy$h`yxhX<tw&meCh`1x1r0guVkP6+0$=b7{V_!
z@A2zKT@~H_tCW!h@;Y&bvpI(qNycbB!bir)ax@BnwEKuW-<jQj2WegU5H<YcuO~US
z+>*}R(;iwquhz65ZRD952A1ExVfM~fi~FLGxR(+iR*_&jTjVDyp5CQn)ESpi*g_$;
z-1oOrT1!yew1b|^I78Jc<YL@*16c1Vcdd9Quo3$b`7L}?JUPC`q>%J9Y@Mx!B7Snb
z7O{qVr_2KT(nE1*U?F=P)}!m1Mq-t6Vd9~rL)t4IUI==u1e5Has+dYjdNUY(A|@s#
z>P}YHw=-EM)+1GHC~{^`OYDDX8*HpSH9NW#U}(GFl9>892jsIs1}R9(g0dX>Bmfh`
zAyT5-j)Du;k5zH9E_?u1C{kFAWJaOi0vlCYD*`9C%KrQC*Ed?G)sLeRWtI0PDABbs
zhm8TOC)VH#yuHEwiooe{o*pI<R^QMbDjpemYK7p3Lh4d7zaNC6r#k3qj6n5(%{og%
zKWHp~TB(erx0yJunduNCTzKa$l+EYf1eJYYuAl{&;MA>h0P`Jtb=gKMs!Y%jd6Z@!
zJLN}x5D7-|HplaK94u1(QD2n!7)|?AE!+D}eUAZH^{jAql6PWMw8cC`hgWu$QtlUV
z951Md#^ZZ>V+3EDb8>c-c<Iz5a47c|iZXo@2I-o!ZPQ~m&gFYqi%NcZ`8yWa`H7-{
z>C4^|XmuIF`vOyJKuxKg#*l?9<pi(N#cxz>wxRX#U!aW*K+(t*UC(qVz_>N7RD#n%
z1L2_790U5MY-yR7#(2{X$5XdgE6;!1ljU%0jyOn0sf9?Ils}Wl8OX}Rj<q2W(#FQd
zGI@zrhn^h+U0fDp^t<0WM5VK0VY`KzrP_9%;|1q5y0$_5)b#iFHx{)G-Y`=I_M-_n
z|0qXCsIp6>Ls@cgiHEyuk|#QZ36!y^NE?;2px|ZEYcM39enVnz*5^s;w}W!!pCZ7%
z#~mbs3hn(j2r4A>w2B`V_(hqeokKcFj^D#n{ZW|eryjPtTh{IogCl)#5?)CGb$arQ
z8j3nq&zB>DKyM5Gfgk?xrSWFGz6_=A29UfM10BqqI-s>QfnA7{LsO=Btcv`_w0KJO
zF+U^Buu8@3;28Du`14mGah@4;*x(lOiDI%U<+g9!prIi>xgPyf=}_!`&MAJ|i5gyn
zh&L4I&pN;V>C^I*31Ln?Eew^C<tnHg(FL2EMVdmK=Lwi#(vmr0t)uh(^{qc#yA3mL
z9hZo{kFw7zt%8=I83QGO)wq`BUbFVfrKr<-FugjCu0tr2y*q!8lEWX^MKxZE*2R5K
zsnVk-N)}t?`Clfo8xX+hgn^768Eqb0VnZ(=kEAZ$xS{=!Oy@d6mv{8NTgkpX|2(=-
zz$tNN_al_K4C}abM8;;aWG(BG*?~rgUYKJr*N1@_P*ZuLpu1ieF3v{Jequs_9(diD
zG|ORCoPX`P4w6=GZtmvV*4EZv`P#KVz?l^HDxorMFgkW@pV)i}LIB%W=k77v29+^a
z?2z7a_CK?%)kUljqRE>&7p3YTk2V5U+W}Im3aAdtU66?8`IOeBRVlIjg>O$}rR>h+
z<e0(^o+^bMH0k*n>7!X8qS@S<n6TueTO5(p-?3o2lOg#K=s!{%J37=|A78{ciEG!H
z6Rk}|t!5v!N!Adx_>rqKd%i{Ctcu4A9uU&A5X7c4(me6q$G*f?E5sHOUO0<EGA<;t
z`!7z&PEJXu9@u98j;3Z7fOZe5u^g7IALlX)$2o+S$RXj%^BKZ^HuyMz;z8w>4i=I!
zJG*vra<WlS_3lxCHT1DY-TbQ7*1dP`-tE*Y>mc=C1zkZx{^F8818dCissrUzd_-q?
zsqN=!$*VOOc5P=GElk661GXaZkanQ<1?ZQ+1495A*}3=@GEsxMBK%V#yB(Q0;dV+F
zbmW5d^~yOwUIuk)ayq*8vtHD)sm532G0LycWR%$-dueQqv=@P(LY~R-y=DQDf1^kG
zT#UM^-VHxXL?_uXlb|&?wh?tyJp0vdj|!xjCfpLTD$``o&0RMmnb3VDC7mqo7En$V
zkrz@-xJYyNkBFiMS21QMJG7?v@tQDk6V9pP4?|F<?4_f<^;FR(ANQjL=V90W47;oA
zdSPD(hJj@r2}{9q6kBx9z=tk0>kp$gZ~7L<9MOyjwyzy&U#JOhN8{xCP=%35UzQVS
zv}2wfs9`daO_{&nrnWo5LIy*Bg=+Msvla?iTX7Gj-5z`Hj$GhciqvYe{G?U%yfYiX
zhNj|WGebvalGd0|F#10&d;LO1rt}Tou5Oy&%Gm5eckIyKGgFV64Lj(S=Diy-dgIIk
zHgwIPxm;T2<#Z|E_Gaxf`F-5rXpUXvHfR3e;Q6apL?S{M03o{r*j?byY_EQDEhm5C
zM}`}0QGNnzAM1RXyWlAtn3k7!?TO4M4T_^{2wHf2@28Xf%Sg=bGZgrFG!T9kiP{L}
zL~J`cCG*m6nY^VUl}B{1!G?&4Wo9iK%#VKvBJ{y-In{pyCzTbo;t3b}b1`;-mIYUl
zz}_ru&{KRVD<tOykuAUXSEs(VU5{yKl1q$Lk9T0f5zK4U0zfw8oLufwq-3#8ai5^M
z4na=++5T>9g60E1{O5Vu)%kz8Uj5UR2sJkwb;W95_~ybPKb9IOxdtsRbV2jB6$u}5
z=BhkhuZwx}0F!s|YK93990Pr*S9%SsYUTZF@N-SkB$m&sx(f%YjW$L$QFZQ*W;vw#
zs)nrrFh^J!#kmwXIuPIcNvo;nSQ$b*Q|>Fx_J^4aD4ZnI*@Y#a1kAp-hQ5+0Jihs|
zDRjM+(Uw0^MEfQ`=bC>mrsq{3fff>NWRVp>3K^h|EVPKd(^=_&YF>j)%m0*|=-ss~
zPhrG4DyZ(PhT8r`G|J!2EhZdKTkL_!t>1j#B+oGAPeVS+`uZF1^ChMji){+;EF1F(
z_MO?X8O!l}xT<g9yWb{}A(vuJYp}%1OxFDcb!gS`;EOFIbiG#!*ZY)xDSEv;raZ?8
zbpy&H=(y)rl!lhCntjM4jP>*G{tw%fY*YCU`Q>?UQ{t%D5TFT>iYleXcjIColXbGk
zCC{%Ro}<v$f4^gOVFg??a(DBld#3Ai51JEy&3cXal8VV=PxGTHBr$M5b&MPHVM3vt
zIyyt@U)IH8omq?pVqe#bp3OC;PRbTr5lCgwJ&RMDblE8T>FJ&=HC#+XL`SkM05)6D
z)z9sE7`RO1B9<625nn8YZm9091%CmKtQ~1*4=I!?^MyvhM?gpM>9wxeXW)I-0M-MQ
zBiB6t1%Ctd!}I76zMGk`gOrLmps$g*FdyzEPv9ilLK8<uN3TGA1BtnxwHi#+ZIle|
zr7hWdL#i?H|5+%7^E_I|MVwZtj+_M<^ku{vkIv_2UC1Kk!U+nvrN~p(t_|1WMfd&J
zrk&l1MoosrkmEVS(_|}c-YdCfxVAi*NuiBxkYmZqHBYxk7xMwl{tK4v7|x!R>Ng+1
zyPn7~R&Ppx4wD&kauhL_TrL;w>)nEh&V*R2^{~Zv6T~98Wc2q_1T*o=@FR`S5%V&U
zP$MErdk5jD<9ZvKu&CBwyHNFzJm_vg=|>B;WVb6M8<l7{U0!VTQDXNewLEf$6MnT9
z$MS_A9^Fn97<O9+JD;o(5RF_B1a0hJ?}BsqLP})jykX;Iye1~n3hVJ785+DH_4}zy
zOE)i~weE%9Z~i4iI*Z88`8dDsh1~vxz4rlvLNQjo?N#2I&CRL47~IUTB?BLsprEpB
zqW)pMbmMAgX&B6jkd+6X=mhiz=a=VlO}(fB4D!MaS*g6}MI>Cf`rO9#1v52W*YRdo
znrkuMNB4e;g9FOPjQ=g{1--r!X$d_!`zExJJ(H|5<(@qPX$z%zuTYq7&0B##QCjK;
z*8$j?Cj1B!z8mr?joZVknHKPizpyv1L2Fo}You$|iV_Jk-$a*nyb1xsNdw|(s&r=Z
z@MOEN^J%jFnHfvzcrL-`4CC6hYt-q_gUb4`lcOQ!F`|c4QBUkJkAB_gJ~AZ|P(Kgs
zB@>xs+w6{akx9^H&n@_`dhJXY9aW{#r~E4uw|*`*&8gl^kbFg$8QwlF`LO@22z`n6
zKzRO!r~OnLeZj3-)&~Pqo>+2ty0XN&jl`5*vAYr`j${y5@`UeFPY=TPIccM(hepTy
zxzwrg{=>6d>`0r=B2yWlFEHz7+bq;ZMtgS&2&>q|u{JzLssJ8enbC^F&UO30<&Ea8
a*@fgacZ&*=ZD#@ex6{ePk@5F{fBrAgYcjV0

diff --git a/lib/matplotlib/tests/test_colorbar.py b/lib/matplotlib/tests/test_colorbar.py
index bb4781f131ed..75e12d3b44cc 100644
--- a/lib/matplotlib/tests/test_colorbar.py
+++ b/lib/matplotlib/tests/test_colorbar.py
@@ -776,11 +776,12 @@ def test_inset_colorbar_layout():
 @image_comparison(['colorbar_twoslope.png'], remove_text=True,
                   style='mpl20')
 def test_twoslope_colorbar():
-    # Note that the first tick = 20, and should be in the middle
+    # Note that the second tick = 20, and should be in the middle
     # of the colorbar (white)
+    # There should be no tick right at the bottom, nor at the top.
     fig, ax = plt.subplots()
 
-    norm = mcolors.TwoSlopeNorm(20, 0, 100)
+    norm = mcolors.TwoSlopeNorm(20, 5, 95)
     pc = ax.pcolormesh(np.arange(1, 11), np.arange(1, 11),
                        np.arange(100).reshape(10, 10),
                        norm=norm, cmap='RdBu_r')
diff --git a/tutorials/colors/colormapnorms.py b/tutorials/colors/colormapnorms.py
index 41aa6310949a..56ad0d948715 100644
--- a/tutorials/colors/colormapnorms.py
+++ b/tutorials/colors/colormapnorms.py
@@ -277,7 +277,8 @@
 # longitude depends on latitude.
 ax.set_aspect(1 / np.cos(np.deg2rad(49)))
 ax.set_title('TwoSlopeNorm(x)')
-fig.colorbar(pcm, shrink=0.6)
+cb = fig.colorbar(pcm, shrink=0.6)
+cb.set_ticks([-500, 0, 1000, 2000, 3000, 4000])
 plt.show()
 
 
@@ -312,7 +313,8 @@ def _inverse(x):
 # ----------------------------------------------------------
 #
 # The `.TwoSlopeNorm` described above makes a useful example for
-# defining your own norm.
+# defining your own norm.  Note for the colorbar to work, you must
+# define an inverse for your norm:
 
 
 class MidpointNormalize(colors.Normalize):
@@ -323,9 +325,18 @@ def __init__(self, vmin=None, vmax=None, vcenter=None, clip=False):
     def __call__(self, value, clip=None):
         # I'm ignoring masked values and all kinds of edge cases to make a
         # simple example...
-        x, y = [self.vmin, self.vcenter, self.vmax], [0, 0.5, 1]
+        # Note also that we must extrapolate linearly beyond vmin/vmax
+        min = self.vmin - (self.vcenter - self.vmin)
+        max = self.vmax + (self.vmax - self.vcenter)
+        x, y = [min, self.vcenter, max], [-0.5, 0.5, 1.5]
         return np.ma.masked_array(np.interp(value, x, y))
 
+    def inverse(self, value):
+        min = self.vmin - (self.vcenter - self.vmin)
+        max = self.vmax + (self.vmax - self.vcenter)
+        y, x = [min, self.vcenter, max], [-0.5, 0.5, 1.5]
+        return np.interp(value, x, y)
+
 
 fig, ax = plt.subplots()
 midnorm = MidpointNormalize(vmin=-500., vcenter=0, vmax=4000)
@@ -334,5 +345,7 @@ def __call__(self, value, clip=None):
                     cmap=terrain_map, shading='auto')
 ax.set_aspect(1 / np.cos(np.deg2rad(49)))
 ax.set_title('Custom norm')
-fig.colorbar(pcm, shrink=0.6, extend='both')
+cb = fig.colorbar(pcm, shrink=0.6, extend='both')
+cb.set_ticks([-500, 0, 1000, 2000, 3000, 4000])
+
 plt.show()

From 38ad08e50c086844bf4e1f829cd2ab694a9c77e5 Mon Sep 17 00:00:00 2001
From: Jody Klymak <jklymak@gmail.com>
Date: Wed, 7 Jul 2021 14:14:03 -0700
Subject: [PATCH 2/2] FIX: use left and right

---
 lib/matplotlib/colors.py          | 15 ++++++---------
 tutorials/colors/colormapnorms.py | 15 ++++++---------
 2 files changed, 12 insertions(+), 18 deletions(-)

diff --git a/lib/matplotlib/colors.py b/lib/matplotlib/colors.py
index 53ac34dfa3db..01005aabf282 100644
--- a/lib/matplotlib/colors.py
+++ b/lib/matplotlib/colors.py
@@ -1300,13 +1300,11 @@ def __call__(self, value, clip=None):
 
         if not self.vmin <= self.vcenter <= self.vmax:
             raise ValueError("vmin, vcenter, vmax must increase monotonically")
-        # must linearly extrapolate for ticks on colorbars that go
-        # beyond vmin and vmax:
-        min = self.vmin - (self.vcenter - self.vmin)
-        max = self.vmax + (self.vmax - self.vcenter)
+        # note that we must extrapolate for tick locators:
         result = np.ma.masked_array(
-            np.interp(result, [min, self.vcenter, max],
-                      [-0.5, 0.5, 1.5]), mask=np.ma.getmask(result))
+            np.interp(result, [self.vmin, self.vcenter, self.vmax],
+                      [0, 0.5, 1], left=-np.inf, right=np.inf),
+            mask=np.ma.getmask(result))
         if is_scalar:
             result = np.atleast_1d(result)[0]
         return result
@@ -1317,9 +1315,8 @@ def inverse(self, value):
         (vmin,), _ = self.process_value(self.vmin)
         (vmax,), _ = self.process_value(self.vmax)
         (vcenter,), _ = self.process_value(self.vcenter)
-        min = vmin - (vcenter - vmin)
-        max = vmax + (vmax - vcenter)
-        result = np.interp(value, [-0.5, 0.5, 1.5], [min, vcenter, max])
+        result = np.interp(value, [0, 0.5, 1], [vmin, vcenter, vmax],
+                           left=-np.inf, right=np.inf)
         return result
 
 
diff --git a/tutorials/colors/colormapnorms.py b/tutorials/colors/colormapnorms.py
index 56ad0d948715..ae3ac8825602 100644
--- a/tutorials/colors/colormapnorms.py
+++ b/tutorials/colors/colormapnorms.py
@@ -325,17 +325,14 @@ def __init__(self, vmin=None, vmax=None, vcenter=None, clip=False):
     def __call__(self, value, clip=None):
         # I'm ignoring masked values and all kinds of edge cases to make a
         # simple example...
-        # Note also that we must extrapolate linearly beyond vmin/vmax
-        min = self.vmin - (self.vcenter - self.vmin)
-        max = self.vmax + (self.vmax - self.vcenter)
-        x, y = [min, self.vcenter, max], [-0.5, 0.5, 1.5]
-        return np.ma.masked_array(np.interp(value, x, y))
+        # Note also that we must extrapolate beyond vmin/vmax
+        x, y = [self.vmin, self.vcenter, self.vmax], [0, 0.5, 1.]
+        return np.ma.masked_array(np.interp(value, x, y,
+                                            left=-np.inf, right=np.inf))
 
     def inverse(self, value):
-        min = self.vmin - (self.vcenter - self.vmin)
-        max = self.vmax + (self.vmax - self.vcenter)
-        y, x = [min, self.vcenter, max], [-0.5, 0.5, 1.5]
-        return np.interp(value, x, y)
+        y, x = [self.vmin, self.vcenter, self.vmax], [0, 0.5, 1]
+        return np.interp(value, x, y, left=-np.inf, right=np.inf)
 
 
 fig, ax = plt.subplots()

<!DOCTYPE html PUBLIC '-//W3C//DTD XHTML 1.0 Transitional//EN' 'http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd'>
<html xmlns='http://www.w3.org/1999/xhtml'>
<head>
<title>pFad - Phonifier reborn</title>
<meta http-equiv='Content-Type' content='text/html; charset=utf-8' />
</head>
<body>
<h1>Pfad - The Proxy pFad of &#169; 2024 Garber Painting. All rights reserved.</h1>


<!-- Disclaimer -->
<p>Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.</p>
<br>
<p>Alternative Proxies:</p><p><a href="http://rainy.clevelandohioweatherforecast.com/php-proxy/index.php?q=https://patch-diff.githubusercontent.com/raw/matplotlib/matplotlib/pull/20582.patch" target="_blank">Alternative Proxy</a></p><p><a href="http://rainy.clevelandohioweatherforecast.com/pFad/index.php?u=https://patch-diff.githubusercontent.com/raw/matplotlib/matplotlib/pull/20582.patch" target="_blank">pFad Proxy</a></p><p><a href="http://rainy.clevelandohioweatherforecast.com/pFad/v3index.php?u=https://patch-diff.githubusercontent.com/raw/matplotlib/matplotlib/pull/20582.patch" target="_blank">pFad v3 Proxy</a></p><p><a href="http://rainy.clevelandohioweatherforecast.com/pFad/v4index.php?u=https://patch-diff.githubusercontent.com/raw/matplotlib/matplotlib/pull/20582.patch" target="_blank">pFad v4 Proxy</a></p></body>
</html>