Closed theAeon closed 7 months ago
Great!!
Also glad it's not too late because think I just found another bug. With the new dimred slot there is problem when using as.liger
from a seurat object that doesn't have any dimreducs:
pbmc <- pbmc3k.SeuratData::pbmc3k.final
pbmc <- UpdateSeuratObject(pbmc)
pbmc$treatment <- sample(c("Treatment1", "Treatment2"), size = ncol(pbmc),
replace = TRUE)
> pbmc_liger <- as.liger(pbmc, datasetVar = "treatment")
Error in dimnames(x) <- dn : 'dimnames' applied to non-array
Ya I think adding to uns makes sense so you don't have coerce it to dense structure.
@mvfki Ignore previous comment about bug. It was error occuring due to rliger2 being loaded at same time as newer rliger.
haha I was about to say I couldn't reproduce it.
haha yup. thought I had removed rliger2 but guess not. Doing that now though lol
Hi @samuel-marsh , it is live on CRAN, surprisingly no much trouble from the manual inspection!
Again, I appreciate it so much that you kindly gave us so much help in testing the package.
Best, Yichen
Awesome congrats!! Happy I could help. I haven't run into any more issues in updating functions so I'll close this issue here and if I run into anything more will open on liger repo. Thanks so much both of you @theAeon and @mvfki for all your help!! The updates are really fantastic, speed was always the main drag for me and the changes are amazing (also having leiden clustering without passing to python which is soooooo slow in Seurat)!!!!
Hi! We're planning on submitting a major rliger upgrade to CRAN in the coming days to weeks. It contains a fairly large number of API changes and (more relevant to your purposes) object structure changes. We will be pushing the updated code to welch-lab/liger@newObj shortly and promoting it to master once CRAN has their way with it. Updated documentation is available at https://mvfki.github.io/liger/. Feel free to reach out if you have any questions!
Best, -Andrew Robbins