Hi DeepST developer, thank you for providing a concise list of package in requirement.txt for the environment. Yet when I was using the exact environment to run the canonical pipeline for single-sample DLPFC dataset, I came across two error regarding the package and environment:
First one was for deepen._get_image_crop(adata, data_name=data_name). Running it and I got this error in return:
module 'PIL.Image' has no attribute 'ANTIALIAS', I solved it by down-versioning Pillow to 9.5.0
Second bug occured during the deepen._fit at line 192 at train.pysc.tl.louvain(cluster_data, resolution = res), it returns an error that no louvain module is found. I am using scanpy==1.9.2 with python-igraph==0.10.4 and python-louvain==0.15. I solved this by installing additional louvain==0.8.1.
Hi DeepST developer, thank you for providing a concise list of package in
requirement.txt
for the environment. Yet when I was using the exact environment to run the canonical pipeline for single-sample DLPFC dataset, I came across two error regarding the package and environment:First one was for
deepen._get_image_crop(adata, data_name=data_name)
. Running it and I got this error in return:module 'PIL.Image' has no attribute 'ANTIALIAS'
, I solved it by down-versioningPillow
to9.5.0
Second bug occured during the
deepen._fit
at line 192 at train.pysc.tl.louvain(cluster_data, resolution = res)
, it returns an error that nolouvain
module is found. I am usingscanpy==1.9.2
withpython-igraph==0.10.4
andpython-louvain==0.15
. I solved this by installing additionallouvain==0.8.1
.