No error when running but does not appear to be converging.
wmorgans opened this issue · 0 comments
wmorgans commented
Hi,
I am currently trying to run scAI. There are no runtime errors (or warnings which would indicate what is happening) but the learned matrix H (from scAI.fit@fit.H) and W contain only very small values. It looks like the method might not be converging.
scAI_outs <- create_scAIobject(raw.data = list(seurat_obj@assays$rna@counts, seurat_obj@assays$atac@counts), do.sparse = T)
scAI_outs <- preprocessing(scAI_outs, assay = NULL)
scAI.fit <- run_scAI(sc_AI_obj, K = 20, nrun = 1, do.fast = T)
Please let me know if you can see an error in the code I've provided or you have an idea what the issue is.
Best,
Will