search

grisslab / scAnnotatR

Other
15 stars 2 forks source link

ambiguous cell types in train_classifier() #9

Closed seb951 closed 1 year ago

seb951 commented 1 year ago

Hi,

train_classifier() considers cell types containing "/", ",", "-", "+", ".", "and", "or", "(", ")", as "ambiguous" . Perhaps this is a fine default, but it would be nice if this was an argument one could set when calling train_classifier(). Otherwise we have to recode cell types which can be totally valid and unique in a dataset...

Thanks.

jgriss commented 1 year ago

Hi @seb951 ,

Thanks a lot for the feedback! I've created a patch already (see above). You can directly install the updated R package from our develop branch using devtools:

devtools::install_github(grisslab/scAnnotatR[develop])

@nttvy If you are OK with the changes, I'd merge them into main and push to BioConductor.

nttvy commented 1 year ago

Hi @jgriss

Perfect! Thanks a lot! I'm OK with the changes.