Can I use preprocessed logcounts as input data? I tried setting normalize_library and log as False in ltv.utils.standard_clean_recipe(), it still preprocess the spliced and unspliced layers. Any suggestions?
Another problem is, after setting normalize_library and log as False in ltv.utils.standard_clean_recipe() and running train_vae(), it raises
KeyError: 'spliced_size_factor'
Hi,
Can I use preprocessed logcounts as input data? I tried setting normalize_library and log as False in ltv.utils.standard_clean_recipe(), it still preprocess the spliced and unspliced layers. Any suggestions?
Thanks.