sometimes need to explicitly cast to character

RGLab · May 28, 2015 · b0aa5c8 · b0aa5c8
1 parent 019bdc6
commit b0aa5c8
Show file tree

Hide file tree

Showing 2 changed files with 4 additions and 3 deletions.
diff --git a/DESCRIPTION b/DESCRIPTION
@@ -1,8 +1,8 @@
 Package: MAST
 Type: Package
 Title: Model-based Analysis of Single Cell Transcriptomics
-Version: 0.930
-Date: 2015-05-26
+Version: 0.931
+Date: 2015-05-28
 Author: Andrew McDavid <amcdavid@fredhutch.org>, Greg Finak
     <gfinak@fredhutch.org>, Masanao Yajima <myajima@fredhutch.org>
 Maintainer: Greg Finak <gfinak@fredhutch.org>

diff --git a/R/ZlmFit.R b/R/ZlmFit.R
@@ -232,7 +232,8 @@ setMethod('summary', signature=c(object='ZlmFit'), function(object, logFC=TRUE,
         names(llrt) <-  doLRT
         llrt <- data.table(melt(llrt))
         setnames(llrt, c('test.type', 'L1', 'value'), c('component', 'contrast', 'Pr(>Chisq)'))
-        llrt[,component := c(disc='D', cont='C', hurdle='H')[component]]
+        llrt[,':='(component=c(disc='D', cont='C', hurdle='H')[component],
+                   primerid=as.character(primerid))]
         setkey(llrt, primerid, component, contrast)
         setkey(dt, primerid, component, contrast)
         dt <- merge(llrt, dt, all.x=TRUE, all.y=TRUE)