search

satijalab / sctransform

R package for modeling single cell UMI expression data using regularized negative binomial regression
GNU General Public License v3.0
213 stars 33 forks source link

Consider standard lapply when there is no parallel processing #201

Open jgarthur opened 6 days ago

jgarthur commented 6 days ago

Since future_lapply is used in get_model_pars even when n_workers == 1, the user needs to adjust future.globals.maxSize depending on the dataset, which is tricky for incorporating sctransform into a pipeline. I would also worry that future_lapply is unnecessarily copying a lot of memory. Could the standard lapply be substituted instead when there is only 1 worker?

saketkc commented 5 days ago

Thanks Joey! I have been myself hitting this lately - will fix it!