From 120cc15f352d59c80461aca601036d27f994f469 Mon Sep 17 00:00:00 2001
From: Gabriel Hoffman <gabriel.hoffman@gmail.com>
Date: Wed, 20 Nov 2024 14:14:22 -0500
Subject: [PATCH] update docs

---
 TODO                         |   3 +++
 docs/articles/cell_covs.html |  28 ++++++++++++++++++++++++----
 docs/favicon.ico             | Bin 15086 -> 0 bytes
 vignettes/cell_covs.Rmd      |   7 +++++++
 4 files changed, 34 insertions(+), 4 deletions(-)
 delete mode 100644 docs/favicon.ico

diff --git a/TODO b/TODO
index 466d994e..6774eec3 100644
--- a/TODO
+++ b/TODO
@@ -1,4 +1,7 @@
 
+Nov 11, 2024
+# Better memory usage for parallel runs with SharedObject
+	https://www.bioconductor.org/packages/release/bioc/html/SharedObject.html
 
 
 
diff --git a/docs/articles/cell_covs.html b/docs/articles/cell_covs.html
index 851a52b9..f4d47fa1 100644
--- a/docs/articles/cell_covs.html
+++ b/docs/articles/cell_covs.html
@@ -26,7 +26,7 @@
 
     <a class="navbar-brand me-2" href="../index.html">dreamlet</a>
 
-    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">1.4.0</small>
+    <small class="nav-text text-muted me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="">1.4.1</small>
 
 
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
@@ -72,8 +72,8 @@ <h3 data-toc-skip class="subtitle">Collapse using mean value
                         <h4 data-toc-skip class="author">Developed by <a href="http://gabrielhoffman.github.io/" class="external-link">Gabriel Hoffman</a>
 </h4>
             
-            <h4 data-toc-skip class="date">Run on 2024-11-05
-15:35:04</h4>
+            <h4 data-toc-skip class="date">Run on 2024-11-20
+14:13:13</h4>
       
       <small class="dont-index">Source: <a href="https://github.com/DiseaseNeurogenomics/dreamlet/blob/HEAD/vignettes/cell_covs.Rmd" class="external-link"><code>vignettes/cell_covs.Rmd</code></a></small>
       <div class="d-none name"><code>cell_covs.Rmd</code></div>
@@ -122,7 +122,7 @@ <h2 id="standard-processing">Standard processing<a class="anchor" aria-label="an
 <span></span>
 <span><span class="co"># only keep singlet cells with sufficient reads</span></span>
 <span><span class="va">sce</span> <span class="op">&lt;-</span> <span class="va">sce</span><span class="op">[</span><span class="fu"><a href="https://rdrr.io/r/base/colSums.html" class="external-link">rowSums</a></span><span class="op">(</span><span class="fu"><a href="https://rdrr.io/pkg/BiocGenerics/man/dge.html" class="external-link">counts</a></span><span class="op">(</span><span class="va">sce</span><span class="op">)</span> <span class="op">&gt;</span> <span class="fl">0</span><span class="op">)</span> <span class="op">&gt;</span> <span class="fl">0</span>, <span class="op">]</span></span>
-<span><span class="va">sce</span> <span class="op">&lt;-</span> <span class="va">sce</span><span class="op">[</span>, <span class="fu"><a href="https://rdrr.io/pkg/SummarizedExperiment/man/SummarizedExperiment-class.html" class="external-link">colData</a></span><span class="op">(</span><span class="va">sce</span><span class="op">)</span><span class="op">$</span><span class="va">multiplets</span> <span class="op">==</span> <span class="st">"singlet"</span><span class="op">]</span></span>
+<span><span class="va">sce</span> <span class="op">&lt;-</span> <span class="va">sce</span><span class="op">[</span>, <span class="fu">colData</span><span class="op">(</span><span class="va">sce</span><span class="op">)</span><span class="op">$</span><span class="va">multiplets</span> <span class="op">==</span> <span class="st">"singlet"</span><span class="op">]</span></span>
 <span></span>
 <span><span class="co"># compute QC metrics</span></span>
 <span><span class="va">qc</span> <span class="op">&lt;-</span> <span class="fu"><a href="https://rdrr.io/pkg/scuttle/man/perCellQCMetrics.html" class="external-link">perCellQCMetrics</a></span><span class="op">(</span><span class="va">sce</span><span class="op">)</span></span>
@@ -211,6 +211,26 @@ <h2 id="analysis">Analysis<a class="anchor" aria-label="anchor" href="#analysis"
 <span><span class="co">## coefNames(4): (Intercept) StimStatusstim value1 value2</span></span></code></pre>
 </div>
 <div class="section level2">
+<h2 id="details">Details<a class="anchor" aria-label="anchor" href="#details"></a>
+</h2>
+<p>A variable in <code>colData(sce)</code> is handled according to if
+the variable is</p>
+<ul>
+<li>continuous: the mean per donor/cell type is stored in
+<code>metadata(pb)$aggr_means</code>
+</li>
+<li>discrete
+<ul>
+<li>[constant within each donor/cell type] it is stored in
+<code>colData(pb)</code>
+</li>
+<li>[varies within each donor/cell type] there is no good way to
+summarize it. The variable is dropped.</li>
+</ul>
+</li>
+</ul>
+</div>
+<div class="section level2">
 <h2 id="session-info">Session Info<a class="anchor" aria-label="anchor" href="#session-info"></a>
 </h2>
 <details><pre><code><span><span class="co">## R version 4.3.0 (2023-04-21)</span></span>
diff --git a/docs/favicon.ico b/docs/favicon.ico
deleted file mode 100644
index 27a5169a52629e6f31bcbf09f228ddefae9f5a37..0000000000000000000000000000000000000000
GIT binary patch
literal 0
HcmV?d00001

literal 15086
zcmeHOSx}tWb;jjXWnMDn%9E<*Wge2MJo+h>hdek=oTZYgY+e%E6WbF{JXN;Gvv@o+
z+K17KBqR_>fRKbhh`m890t8}5Y?43{&|+zpX6>f;egE>E+uhJCV$(SGLoW5x>woV#
z_n!4Tr<0PtlJwQ2uYWy>k8dUw|7lXvpC=_Hz4>PB_1BV;rug31_$Z&h`v*x$r}%|G
z<9B#V(vNw``)d42^kJ+20AAzO%<I)hB#eM#8qTS5ICZ7)+b0o;gkF8#=nOr4_@j+H
z@VzgE{*ccM&tf}VlSS~&RKho{fqU`}+zYJ;c~`hhBud5p|NmC#4+q`wt~`NDR|fA?
z83GH<2+Y;PKU)Pq`S9vW$WsZt#(snYuKiOFe5v$D!a?|L<8Vz^!ZT5bz*0NH?s?MR
zr05pf*Deh_JPQ}k#BbFi5)SN#k1vP*NCZLmBHVM0N<RGhd))5^`1N(<p^f`qoj`XY
z4<b|Lyr&RueLVuMIf_(xFMPZ*{UN^%o~3Rj<02Qb=Ml<JgzpFDx$n~zFQt9$EXc=H
zDd)IDed$2RXW1<e`=&n<^1^3&s$|@wD<X}Q1Ij{Rz7b*PEJBVcp259c(VobMS69qA
zOO^hFgPxuE*fafN`vUeUxMr%OvJ&6l*}L5P4g^;Q5Lj#>-BDdlq;1#Fxerm@;2gFI
zid67reC&aK%DBe>_k6RW+ef>yMvn$T(_<jEP9)?}yvTW|bCI)Iz2fqp_+H}k_~!|_
zLA?_FqT|@e$FArP2ORJkdMTg9%J!|%9bZd>kmWh&j?GV75;XQwe`_gE_oMccx*#?*
zpsz-Nb~LoqjEMCqB2FFUO^1+m7{TQ(bx)Mn_}sL&N;is>o{7FCjBV+Ug#7SXM&O#N
z;CaSme4PzkTSIJ2&@_m!*N6!9FU+%w1niKrh=hEg!$Q#Ogw?PNn|T(^%_ej|@uG3e
z1G6m*uQ!5lI7GcSBW!*gvsrR3@iiz@)HCrZJ~R2|d>k*OpLV?2O*<(5(ifNU)w$wx
z$|KjH><0}GY3q5$VSmE*tbh9Wu@7IIupp)0ieIG|QCjOjPKOPbGOVb+>%xs2b_@>s
zF+1x=n6^}G!N#)?dnPowW^3p-Ui5^6-c9kt^5fI5*+#OO<=oczPoz2ioc4w?5vHw)
z%YGuo9!~^TYZ$?B1kV<IICaj9k`f2bW}0y--Hh|+Ex4L#!<Fk+6m-~e?WPSyMGoZU
z+0`@=#uy@hn|R<pivHP0vwIAKDVyk@>w5n+JL!+=U!2yBkEES(L0H8xiBzvW@9Od(
zJ<W>l0S^vVt>D3s8!gRl6ja;sY3T~abbeIzI8iy^#D#kneE1t98X8>4%ChkcqBcj)
zRP<=`K57*Gi#?xiLH`Fk&>yw)5ua&|*e#yTuLO}@Z^P*nGcp^j$nCb_>}fM98lA|0
zY=_kmM#Z2Lw;Jp?mtw)0GiDq+X2SLBHk><WfqqVSh$!2z`3z$8_kOl_`YG??qjnm4
zQxKP1tV%9AdOgt2_!V89LmtIT)-@Y$-n8T7Ni#C21Mj@EjO=Wi(gTggiJ>9tJx7H0
zPsM-A2FK-Kzx31A8~Zrv&M|yvhX+|#Z73+PV`S8i0lgPj+iXa`Y{di0(Wz5rbds;5
zr6yz*@NCXlke_c?Jcu5MK1WTyoWZjVGPF<nX(z*;<(;-(eLlMs#I;%*&QN}Cv~ZvA
znepBU10D@|ai6+S*XV-L974;a8&+=^6B7Z5&a||+vBZ1Drmhb0DW)e0JnV;l%Kc*N
zwl;EgJwBPY1#zX5a-U~Id7TsG^-eV3cSCc}i4O~mSaI!qhKchru%n&sj@hGq(9d{4
z#R=;s-Tp-|6hUjN8x8d?B&V8iqsNYdB0C=TD*v<n&0BmLVNR#wLee7kb8%1l%Q%6#
zh8kh!wd3RPtvrKo<O#wy#=M*3lXVmr>>)Ibx#0^Y$i=#E?CwQ~{#X3izUa4)BjWbJ
zo?igVxwEjfWy4f(9G1pZxaw+XEB2K~?T+x+x5RGy>3{b{KjSRPk!)q8`{-df?^VK9
zeGRs{EZF+5Q8)G|(|e*k_GQK>5~JHkv-T`M!Igf5Y(`j`&%=^?4i?RM*xIwOa`;oY
z`|rck-j0xag?hsHWqJl7qj5hJuDeI1Ex0bPWGv74c0cq77MtOF(hFO5Hk_ShuyyCc
zQF|Na(q!0g<iVW$1+0yku-rQfOXfwmI~v!~vX>XGkPCy3Cz$q_aA9Tu2ETjtcCn><
zr(b*@V|Sa5@o<=VZa`fw`eXC$FL)NIuoh&%_{lF}?M{bF-@trlfAr6K9q6<^$8p^;
zjJR~TGS`G5N6dD}d8|CzC;jvR#1zC9%LrR7aI_S{Qh65EhIG<@5vI=%!JK{?)<>BL
z7^i8QY~=H$IiQ+n<G;Lx=O|nnfz~yO)6=JLY4##YmP?{vqbvoO>+Em;C3d^7@`p$;
z1V{O8SZ<tyt^O*kMHyJB{WUB-7hy?14|57-i85rnmP>t`5pFk<`O?=O+JTZ~4T_hG
zcn<A&;hw|U>255G3`u^<bKNifK?BnsrB7iv$E#qeN`<|+0Ja+HLIuy_Rw^tV=V7{i
z1m?12I5iqLC(D_`65rY7aEIK;n!Ab$Lld45KRq*j5jPfkU=x{TZX@yZ-t8ylUwklS
zPey3P0Q3D*Fl8TxC5!f`hS%A%Fc*CR=Y$59x>T4oCtyiQfvbtWA+melp!4XI>}St+
z;m|}i_dW$n{v{PR2A3H_?8|?}=$H7D!>=8KtDAo0;Z0bsQcmi#V5+1qyG*~<con9u
zL$pH|V985`Q(M71cJFufUOnahG^&mF@Uzih;*;@@G4ESozCTOfCD_a^;s<`mF<#1i
zlV`7w<2e$;wNR9q=f>iNfNmPrTWPS=pM~i%>F0i%(@w+LQ-z>)j<~`sv6ltK^H%@7
z#9jLBOSChSC|W8;lX(&vLkm9FoxsVdI$WHsL@#lmpv0h*-8e3@o%}ep+08S3xj1*^
zZCpxz2P3Ul$OADd;a$bhJj>V!%uT^so(B6sJ_3^y@V|JWu8`Rr|H?}r1C*&{zXKnS
zx8okMuT<KZ<GMy1p1eu9If@K@2R@ysLAw4CLbQ2NY%GcwZAZU*{91Es*-2!b{W&gv
z{tgbl_gz%oIE=YRH`Uom92>_Pf-XJGwP%Pqcf6F&#1BF|H~EuMvRsVYOGP-Q>q3Wh
z9{Gzmada|+^ta)|u|vvselYei=6v+y#D3Ot;qBz-{ns3$S<R;5LY(^Szwq{dehdB7
z#h?H99~D1VZ85Q!3WP0B$wO2p{Uf7Tx8Jl+66@0NCz6gZX}`0iCCy#9y;y`~T{&&b
z74GpF+%?{(y&pmCN)z?<4h~Ho#oHr4L5qd*$`NAhxR%Gq=-(oLUw(W_;V1s-GE`hU
zh{ij~xRv$+-hJyIFw%Af=a2q3j(_xh#Z$yYU6#EOKE$T-G2*w9m#0x|g*aB?kX^4g
zjWZs&P1#PTUe;1J3Xn#*d4KGE#tS3JozJJ-q?3nCv|0M#4Kr6zveT;0UU=9JKOrHn
z#qlfW*lmT;%YJug&Y(K)2)^@Ae~BOd`(NSA!5^cp;3%{mIm827s9Rl>PvZT={S{sy
zwpM&a?2O`kOnN4{w}(i7E9It?vj1y*F!lkej2e_KkK@SXUGk8Nl<7{sH>qw5*{N58
z32(x$Je%+b-k|DV>+nNjcVdIQQyNs{9>n2aeGiq_58-D#lOKKOuTh%y2}IsPt9B}m
z6G*%?#wnxnF2n>WQxon9Wa{sbhjzU8;sR+-#v<eTG4}%H=P`ZP4NURu>>-{f-xvHr
zFieFX61U$8KjbpavX+c>`0W&H6w3ZqVxayaY5yjE`u)FE@^s>px6oI20dC54h_-4~
zR*64_of1E+@i6C`!}*yD^r`LK>ss!03r<bfk&nZ)oA+t=qx&wh=wmHM)izis*$aNF
zh+qA;0Y9g1xG7UZO_y<>{#JN+|DAuvNALbC^3y)#`|0YeqCC*v>1%inFIHuV+Kqz6
z=g4MUnXEg`*m4$=9wXzN7nt^1U<!DMfzHDtb%s~O&sSNH9ZaZSNR1(+ujJm8qc;DT
zvQIhZe!(+H#arM0Ch7}6r|!_-@w@`mIgww%$itKok<BjJ%R$Ey93clIyI5=|K9P3I
zPrY%?Qjb^<{Br6SiCmD6g!+-xd_?Y7o)=>NaUO16`asd%LjQ4ydQx=dHxQc}_2E&x
zT5y<%c?b~)>)DJ61OpCPhT&Uypz4=i>VVW?cNjDduU5Z~*6+G_2CNTIrs|l>i_iA3
zW{{TrKlqow{|mhHgTKSG2N_s?S^)c)=#QKQ`QX{aaZOdfW6jk=`z3YnU5;0#U#-J4
zkJnEfV?v|Ys=LZp9HXB|J^oWXqOOb2$UOBfETcuMHsE(%KTX`Hc>UT<ITzgs%u3Cq
z0!xGW_`$z@12<BCg;DxHp<nj>_ojZG$ONCS^$(QNPnN(ma!18ZqT{v~CF;H3$Nqs0
zY$!RFdbH?9V5XesA+?{P-~0ZJb$m#!t@>vwe(U`+Gtx-&he=8Q|M<Jv2kFy#lh-$R
zt@dfLmuss*8F`&oG!b4WFfVx_azsDno~?zOF@ea`9{aF{IP)LJY2ou-uOIR;hGPBQ
z%eABVj|N`Gv0mZ;d+;@TUSIk*{N`tjffzS4{uJCpYR&PuP;vkddyRaiq22iEU9T^G
zAYj+Q!}_-`7JI7R0HH(0LGjo+>W8IA%Okx)jPbYWW7>Ir@uRG5c^6t$-<RYH(mN$_
zNq})l+}CfWfpo}ukPa`u7wxs!gtcwEe%Nnk8rR4DtWf!qtPzy>h&ivst!hnvwia_R
zE~F2P#`3Z+(O5+MxtfFWvLV*qa((G-@Uurk;#@h;MBapN$@`>^E;UQ3Jqd2CdOH~B
zSZ0P`G3gO>F&=e{A!KNe(W6mwd1fdp=I8NTW!?JWpzHO3ZM@IJ*jn~&{oY1nL|I4W
z;PKoxrkXtu-6(1_BmI^Yg>6pMG`P{%=flL93!asUXnjh_3H3x|Lwd<<V?!aoV^#l*
zgXdm%j@{~$b4%n|=#cuB>Pbk*{Y|D2rgQ=H>b$sl$Btaq(9YzTai@$u6jiLr3N3{W
z==dE9K`e)j$tl-xjo%tt8M?5#{&L@u*ZE%G#<eB)mH2++m|bKozKH8+thm$6+HQ#*
zwa;DX81>-VT^mlEH6!g3KfB|Ajr}NLdrW3x`r;bTA2JMlzP5h<(Jj_zu1btzt1-JA
zL~f-G4b3hz^}A76<v?*G>0qx#YL*2@j<5#Jz6DwT=~H5JVPeG8o%p$`^&cf~vp)Bo
zxSr~@*@A%S`2eow+fdTrz-9I#+~)dKO-}SY@}lL18v}f-XN^<p$5K0%Aw4WnEP*+e
z@V~?QO5V3vTi&w!!l>(WA&+ODS7AqSwF7tAyRp*<v0qp4&Yjn<R6Pw5e~_35YgbDX
ze9b-<$>(At5Qt#K7(&}~4~C|EaJ$w%i+^X+BXuXZ+^OVmd;YmTdqq8MZLnq$3%<vC
zV0|`encp<!Uf&2ibW#3ykiQUfRcChv%-7Da9?5#foiuod2M}`7{wywVKX%`1D>$7$
zR?8F{!nk$F(0_D$`J3y2=V2#2BkgeZmcdeV0oII5uvD<V(r}$Njd<Y>gRM%$k+sB&
zpP^=DOw|pgkIbzr9WwMC-e!H;3i>)_?^s)NKWk=PBn6g|3ou<c3R_1m{KOq2-e}$G
zWn>IG48*+ph^v<3`a&TL0rvQ?pKWLUxxVz(c$ewtmxxDaur8d-njUMccUi~CJ^^c9
zCcMud(#O8Mcfih=p_p}}4ofE=+wq9_1c)8&%Kixd+$;b5FzcbLT@|t(cQc*qUxc+S
z9p+mn;A|=*p7nCQTk<aw)47Gq`bOl;--46<*@9Jz{ar8rQTr?YQ*a4|uMi`W{SWA;
z;eOP>dRiLHB_~<iVQsMW66`GnJeNiRKh|x()P|j`iJ1dd;=hxuRbD5ilz~cP4Sl23
zJMKpLPq05LkB+}?8@<`rR&g2+N>9Shyin*6oKi+`X`VWG6|T-|cv^{xYDW}~vk}V)
z5&!SEFA$UNC1!h@7+eKTOkLsn_gJ@^MTC8sVuLr@AL~%^>&B-!eb+C%hvbich&*C8
z!>ySJFSl@UCrf-5>Huwm&k^NSu;m2&&lK<yue7jM(oXESh;@?zlp89EXIA2qiDSf$
zFTmhiK}dQ!<Lf8rzxZFzWI<nR`AL)zJ3W2yM=D1yy7Dn*hcuK~_FsB<&cwL{{|uT3
z84E@G`h=h!o1VD4)C!7NBgjO`^f|OzdWdCb^Zg7wbB@ptvrmFDB|d(m|7G7k<By3x
z=S)3Z6UCT(m;=rAgLwNtzJ<(_@ACN(7`4@;zm@VJb4c&aAh#eI*RA-C)UPspF;zqS
z`KYSxRvAZl1_dZv9)dmKX72%GcJ@GS6@QQg75~TMFCS&kIbMwMu51)t{SAKdy}uz2
z^%L~go+o`}QS6L$Bk?D3pU`iYdXI+u77>SSA)Z--GQ$)2h|L96+|IuFm&9N3^<^(*
zMB~qX%4#KXyaKiEr@#0%@w@M$Eay|=UzZgo91tAU%Dv?pb3QxkH9crHcMvl*V}@~$
z(Pe<2*qzVN$D^}GCy)ML7yt3C&G}C({`N7plsI;(CkI`3Q&5=kYn11Frs&YK{uG*H
zzw6=yc*&!kz2MT17GVua&c`ots_J##Tl}3!XgvS$FEn7RGZVe6>$bCYA((7!{!xXS
z3)U|-LoO9_%WoDG)*7uRCC(RnzBB(DZi?q$KH3k#^h7_eo%^M_rhDlRh{^mm@~;hK
zFa?nOvxmRo(GXWpgSP$3|4IJ2j*eLV?VGEEmooS&`FC4V(pP_%m*5lvOT5Wz6{p~D
z%{C75e}EU0u!O;V2vNS3j<Bab#2Q81CMA4J8fD(#weEh=@c?sm={<?p*nG4pe&+Sz
z^?mj0?iZ{-z`i}7U;v_D(%VnD57Jfy+0QRnv!CY@u#U!aJ#|0VI}mi}hW)fns)t?b
z-qcg+vG?2c&^{f<!@gOJYZqaiq3&`n!T*DHotC-l8#e8=1GJyRDu)oeqikOgcJ>`V
z==UK1t`h}y4rJW2p}N`$w_!}_Bk$4Lw7G9spQj(-Z*tWAQV)&$FQX-dJM4WgtY-}W
zz=?`lCu(b5Ft2!Nv%BRt4qEH*W!`@|dcV+ZpJyI!qAwyQ!t`Qr)Q6^CHx}6=&2SAt
z_9RL)uIgR-I<56->gxT0xmx(N-LO^Wut%+lbD34&6k9XbjUeNSAY<0R%44nN$*I-*
zWv>HzVgt|GVJ*4DKA9Yzfqvb+4(90{)<<~moI=ntthK89(+{vNF)a0p5U~XJlLxRA
zv$v)+1<sZ-##(D}oH@YW`K20knThc%b!x4HXWsCRXC2V@=M3uyZYb<0pdW+1?xu=K
z{CbKdaT=ldg=>*D<!;sk3ovTp?{8)*wbtipZ@^G;psy}v`1a)w(OsJc>)cb)F+;!3
z-~KT#F)=pN5Hq+sSHgOy#H;MdC&u8@MbD3M$N`^qV%Yei2F)d3ATRAB^s-0ArK>`O
z`K|P~-CV3k##}M`d^#~roFquxBWRyg^1}^^%756WFqEl*+w>!!ur_rZ6A!N|EJH_(
zV~YOFMB5!?z8H!AW@(-LuWoeE0mniwMp>^eS3WVF=W<W^T8UfZ^2H}x->)X(??k*y
cj4kfMpk=%#xqd9BSY0P}|M9ORCB4e^ztgSgEC2ui

diff --git a/vignettes/cell_covs.Rmd b/vignettes/cell_covs.Rmd
index fb8833e3..14c2c84e 100644
--- a/vignettes/cell_covs.Rmd
+++ b/vignettes/cell_covs.Rmd
@@ -129,6 +129,13 @@ res.dl <- dreamlet(res.proc, form)
 res.dl
 ```
 
+# Details
+ A variable in `colData(sce)` is handled according to if the variable is
+
+-  continuous: the mean per donor/cell type is stored in `metadata(pb)$aggr_means`
+- discrete
+  - [constant within each donor/cell type] it is stored in `colData(pb)`
+  - [varies within each donor/cell type] there is no good way to summarize it.  The variable is dropped.