global stats

evaluate.py

@@ -12,6 +12,8 @@ def parse_args():
     parser = argparse.ArgumentParser(description='Process some integers.')
     parser.add_argument('filename', type=str,
                         help='The output file to evalute')
+    parser.add_argument('--light-print', '-l', action='store_true',
+                        help='Print only wrong nodes and paths')
 
     args = parser.parse_args()
 
@@ -78,14 +80,35 @@ def parse_graph_frequencies(graph, only_wrong=False, file_pointer=sys.stdout):
 
 
 def print_summary(frequencies, file_pointer=sys.stdout):
+    if file_pointer == None:
+        return
+
     print("--- Global summary ---", file=file_pointer)
 
+    # --- Frequency usage ---
+    # Tags
+    distinct_theoritical_nodes = len(frequencies[0])
+    distinct_observed_nodes = len(frequencies[1])
+    print(f"Distinct barcodes: {distinct_observed_nodes}/{distinct_theoritical_nodes}", file=file_pointer)
+    # molecules
+    cumulative_theoritical_nodes = sum(frequencies[0].values())
+    cumulative_observed_nodes = sum(frequencies[1].values())
+    print(f"Molecules: {cumulative_observed_nodes}/{cumulative_theoritical_nodes}", file=file_pointer)
+    # Wrong splits
+    over_split = 0
+    under_split = 0
+    for barcode in frequencies[0]:
+        observed = frequencies[1][barcode]
+        theoritic = frequencies[0][barcode]
+        over_split += max(observed-theoritic, 0)
+        under_split += max(theoritic-observed, 0)
+    print(f"Under/Over splitting: {under_split} - {over_split}")
 
 
 def main():
     args = parse_args()
     graph = load_graph(args.filename)
-    frequencies = parse_graph_frequencies(graph)
+    frequencies = parse_graph_frequencies(graph, only_wrong=args.light_print)
 
     print_summary(frequencies)