[Bioc-devel] fData(x) <- gives errors for ExpressionSet objects

Thanks Gordon for the report.

I think the row names being objected to are the varMetadata row names
(only two of them), rather than the pData row names. More importantly,
I think this was fixed in Biobase 1.16.1, on October 15 (reported by
Richard Pearson around that time).

Sorry for the inconvenience; if there are other problems, or this is
not addressed in the patch, then please let me know.

Martin

Gordon Smyth <smyth at wehi.EDU.AU> writes:

I've a lot trouble adding information to the featureData of
ExpressionSet objects. Here is a small reproducible example which
shows up one error message.

 > library(Biobase)
 > example(ExpressionSet)

 > fData(expressionSet)

NULL data frame with 500 rows

 > df <- data.frame(a=1:500,b=1:500)
 > rownames(df) <- featureNames(expressionSet)
 > eS <- expressionSet;
 > all.equal(row.names(fData(eS)),row.names(df))

[1] TRUE

 > fData(eS) <- df

This all seems good so far. I've been able to overwrite the pData()
data.frame. But suppose I add a column first. It shouldn't make any
difference, but now replacing the data.frame now longer works. The
error complains about row.names, but these are still apparently correct.

Gordon

 > eS <- expressionSet
 > fData(eS)$Number <- 1:500
 > all.equal(row.names(fData(eS)),row.names(df))

[1] TRUE

 > fData(eS) <- df

Error in `row.names<-.data.frame`(`*tmp*`, value = c("a", "b")) :
   invalid 'row.names' length

 > sessionInfo()

R version 2.6.0 (2007-10-03)
i386-pc-mingw32

locale:

LC_COLLATE=English_Australia.1252;LC_CTYPE=English_Australia.1252;LC_MONETARY=English_Australia.1252;LC_NUMERIC=C;LC_TIME=English_Australia.1252

attached base packages:
[1] tools     stats     graphics  grDevices
utils     datasets  methods   base

other attached packages:
[1] Biobase_1.16.0

--
Martin Morgan
Computational Biology Shared Resource Director
Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N.
PO Box 19024 Seattle, WA 98109

Location: Arnold Building M2 B169
Phone: (206) 667-2793