I'm trying to run SC3 on a dataset that was generated with the read10XCounts() function from DropletUtils. In this SingleCellExperiment object, counts() and logcounts() are sparse matrices (dgCMatrix class). SC3 throws some errors when I run it on this object. I think are related to the sparse matrix, because if I convert counts() and logcounts() to a normal matrix, it runs without issues.
Here the details:
The first error I get is during the gene filtering step:
sce = sc3(sce ks = 5:8, biology = TRUE, n_cores = 8, gene_filter = T)
> traceback()
4: stop("'x' must be an array of at least two dimensions")
3: rowSums(counts(object) == 0)
2: sc3_prepare(sce_info)
1: sc3_prepare(sce_info)
If I skip this step by setting gene_filter=F, I instead get a different error:
Hello,
I'm trying to run SC3 on a dataset that was generated with the read10XCounts() function from DropletUtils. In this SingleCellExperiment object, counts() and logcounts() are sparse matrices (dgCMatrix class). SC3 throws some errors when I run it on this object. I think are related to the sparse matrix, because if I convert counts() and logcounts() to a normal matrix, it runs without issues.
Here the details:
The first error I get is during the gene filtering step:
If I skip this step by setting
gene_filter=F
, I instead get a different error:Session info