search

welch-lab / liger

R package for integrating and analyzing multiple single-cell datasets
GNU General Public License v3.0
389 stars 78 forks source link

Problem with subsetLiger #125

Closed JoseMMuino closed 4 years ago

JoseMMuino commented 5 years ago

Dear MacosKolab,

Thanks for the amazing tool that you have developed.

When I run subsetLiger I get the error "Error in .rowNamesDF<-(x, value = value) : invalid 'row.names' length"

The problem is that when I try to subset my liger object, one of the samples only have one cell. This means that it will be not a "dgCMatrix" object anymore but a vector. I see that you eliminate samples that have not cells with missing <- sapply(raw.data, is.null) raw.data <- raw.data[!missing]

I think that it will better to also eliminate elements that are not a dgCMatrix object

Best, Jose

jw156605 commented 4 years ago

I think that this may be a feature rather than a bug, since it doesn't make sense to perform subsequent analyses (such as iNMF factorization) when one dataset contains only a single cell. Perhaps a better solution would be to remove all cells from that dataset, or use different subset parameters so that more than one cell is left.