Merge branch 'master' into pc_revamp

Author: pcahan1

README.md

@@ -1,7 +1,6 @@
 [![Documentation Status](https://readthedocs.org/projects/pysinglecellnet/badge/?version=latest)](https://pysinglecellnet.readthedocs.io/en/latest/?badge=latest)
 
 # PySingleCellNet: classify scRNAseq data in Python
-
 PySingleCellNet (PySCN) predicts the 'cell type' of query scRNA-seq data by Random forest multi-class classification. See [Tan & Cahan 2019] for more details. PySCN includes functionality to aid in the analysis of engineered cell populations (i.e. cells derived via directed differentiation of pluripotent stem cells or via direct conversion).
 
 [Tan & Cahan 2019]: https://doi.org/10.1016/j.cels.2019.06.004

pySingleCellNet/scn_train.py

@@ -153,7 +153,10 @@ def scn_train(aTrain,dLevel,nTopGenes = 100,nTopGenePairs = 100,nRand = 100, nTr
 
         print("HVG")
         if limitToHVG:
-            sc.pp.highly_variable_genes(adNorm, min_mean=0.0125, max_mean=6, min_disp=0.25)
+            try:
+                sc.pp.highly_variable_genes(adNorm, min_mean=0.0125, max_mean=4, min_disp=0.5)
+            except Exception as e:
+                raise ValueError(f"PySCN encountered an error when selecting variable genes. This may be avoided if you do not call scale or regress_out on the training data. Original error text: {repr(e)}") 
             adNorm = adNorm[:, adNorm.var.highly_variable]
 
         sc.pp.scale(adNorm, max_value=scaleMax)
@@ -204,7 +207,7 @@ def check_adX(adata: AnnData) -> AnnData:
 def add_classRes(adata: AnnData, adClassRes, copy=False) -> AnnData:
     cNames = adClassRes.var_names
     for cname in cNames:
-        adata.obs[cname] = adClassRes[:,cname].X
+        adata.obs[cname] = adClassRes[:,cname].X.toarray()
     # adata.obs['category'] = adClassRes.obs['category']
     adata.obs['SCN_class'] = adClassRes.obs['SCN_class']
     return adata if copy else None