On Wed, Mar 12, 2014 at 3:45 PM, Martin Morgan <mtmor...@fhcrc.org> wrote:
> On 03/12/2014 03:02 PM, Wolfgang Huber wrote:
>
>> Hi Martin, Mike
>>
>> a DESeq2 user brought up the observation that when he subsets a
>> 'DESeqDataSet' object (the class inherits from 'SummarizedExperiment') by
>> samples, he often ends up with unused factor levels in the colData. (Esp.
>> since the subsetting is often to select certain subgroups). Would either of
>> the following two make sense:
>>
>> - a 'droplevels' method for 'SummarizedExperiment' that efficiently and
>> conveniently removes unused levels, i.e.
>> x = x[, x$tissue %in% c("guts", "brains")]
>> x = droplevels(x)
>>
>
> vs. x$tissue = droplevels(x$tissue)
>
>
>
This seems preferable.
> - a 'droplevels' argument (default: FALSE)
>> x = x[, x$tissue %in% c("guts", "brains"), droplevels=TRUE]
>>
>
> there are a surprising number of places were levels could be dropped --
> each column of colData, each column of (possibly two levels of) 'mcols' on
> the row data, and the seqlevels of the row data.
>
> Does this make sense lower in the class hierarchy, e.g., Vector, as well
> as GRanges/List?
>
>
For something fundamental like Rle, it's probably best/simplest to just
call droplevels explicitly after extraction. For more complex data
structures, I'm not sure there is really a use-case where the user would
want to drop levels on everything. Probably best to be explicit.
> Martin
>
>
>
>> Wolfgang
>>
>> _______________________________________________
>> Bioc-devel@r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/bioc-devel
>>
>>
>
> --
> Computational Biology / Fred Hutchinson Cancer Research Center
> 1100 Fairview Ave. N.
> PO Box 19024 Seattle, WA 98109
>
> Location: Arnold Building M1 B861
> Phone: (206) 667-2793
>
>
> _______________________________________________
> Bioc-devel@r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/bioc-devel
>
[[alternative HTML version deleted]]
_______________________________________________
Bioc-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/bioc-devel
