add log info

panpipes/python_scripts/rerun_find_neighbors_for_clustering.py

@@ -35,20 +35,19 @@
     neighbor_dict = read_yaml(args.neighbor_dict) 
 
 sc.settings.n_jobs = int(args.n_threads)
-L.info("Running with options: %s", args)
 
 # read data
-L.info("reading mudata")
+L.info("Reading in MuData from '%s'" % args.infile)
 mdata = read(args.infile)
 
 
 
 for mod in neighbor_dict.keys():
     if neighbor_dict[mod]['use_existing']:
-        L.info('using existing neighbors graph for %s' % mod)
+        L.info('Using existing neighbors graph for %s' % mod)
         pass
     else:
-        L.info("computing new neighbors for %s" % mod)
+        L.info("Computing new neighbors for %s" % mod)
         if type(mdata) is MuData:
             adata=mdata[mod]
         if (neighbor_dict[mod]['dim_red'] == "X_pca") and ("X_pca" not in adata.obsm.keys()):
@@ -82,6 +81,6 @@
 
 
 
-L.info("saving data")
+L.info("Saving updated MuData to '%s'" % args.outfile)
 mdata.write(args.outfile)
-L.info("done")
+L.info("Done")

panpipes/python_scripts/run_clustering.py

@@ -33,6 +33,7 @@
 L.info(args)
 
 # read data
+L.info("Reading in data from '%s'" % args.infile)
 mdata = mu.read(args.infile)
 if type(mdata) is AnnData:
     adata = mdata
@@ -45,23 +46,30 @@
 # check sc.pp.neihgbours has been run
 if uns_key not in adata.uns.keys():
     # sys.exit("Error: sc.pp.neighbours has not been run on this object")
+    L.warning("Running neighbors with default parameters since no neighbors graph found in this data object")
     sc.pp.neighbors(adata)
     uns_key="neighbors"
 
 
 # run command
 if args.algorithm == "louvain":
+    L.info("Running Louvain clustering")
     sc.tl.louvain(adata, resolution=float(args.resolution), key_added='clusters', neighbors_key=uns_key)
 elif args.algorithm == "leiden":
+    L.info("Running Leiden clustering")
     sc.tl.leiden(adata, resolution=float(args.resolution), key_added='clusters', neighbors_key=uns_key)
 else:
-    sys.exit("algorithm not found: please specify 'louvain' or 'leiden'")
+    L.error("Could not find clustering algorithm '%s'. Please specify 'louvain' or 'leiden'" % args.algorithm)
+    sys.exit("Could not find clustering algorithm '%s'. Please specify 'louvain' or 'leiden'"  % args.algorithm)
 
 #mdata.update()
 ## write out clusters as text file
+L.info("Saving cluster column to csv file '%s'" % args.outfile)
 clusters = pd.DataFrame(adata.obs['clusters'])
 clusters.to_csv(args.outfile, sep='\t')
 
+L.info("Saving cell numbers per cluster to csv file")
+
 tmp = clusters['clusters'].value_counts().to_frame("cell_num").reset_index().rename(columns={"index":"cluster"})
 tmp.to_csv(os.path.dirname(args.outfile) + "/cellnum_per_cluster.csv")
 

panpipes/python_scripts/run_umap.py

@@ -39,6 +39,7 @@
 L.info(args)
 
 # read data
+L.info("Reading in data from '%s'" % args.infile)
 mdata = mu.read(args.infile)
 if type(mdata) is AnnData:
     adata = mdata
@@ -55,7 +56,7 @@
 # check sc.pp.neihgbours has been run
 if uns_key not in adata.uns.keys():
     # sys.exit("Error: sc.pp.neighbours has not been run on this object")
-    L.warning("running neighbors with default parameters since no neighbors graph found in this data object")
+    L.warning("Running neighbors with default parameters since no neighbors graph found in this data object")
     sc.pp.neighbors(adata)
     uns_key="neighbors"
 
@@ -68,6 +69,7 @@
 
 
 # what parameters?
+L.info("Running UMAP on neighbors_key %s" % uns_key)
 if uns_key =="wnn":
     mu.tl.umap(adata, min_dist=float(args.min_dist), neighbors_key=uns_key)
 else:
@@ -81,4 +83,5 @@
 
 # save coordinates to file
 # (note this saves values values up to 6 significant figures, because why save 20 for a plot
+L.info("Saving UMAP coordinates to csv file '%s'" % args.outfile)
 umap_coords.to_csv(args.outfile, sep = '\t')