add a method to start writing the qc files again

monarch-initiative · Aug 25, 2023 · a87ab02 · a87ab02
1 parent 3e5163c
commit a87ab02
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 0 deletions.
diff --git a/cat_merge/file_utils.py b/cat_merge/file_utils.py
@@ -237,3 +237,10 @@ def write(kg: MergedKG, name: str, output_dir: str):
     write_df(df=kg.edges, filename=edges_path)
 
     write_tar(tar_path, [nodes_path, edges_path])
+
+def write_qc(qc: MergeQC, name: str, output_dir: str):
+    duplicate_nodes_path = f"{output_dir}/qc/{name}-duplicate-nodes.tsv.gz"
+    dangling_edges_path = f"{output_dir}/qc/{name}-dangling-edges.tsv.gz"
+
+    write_df(df=qc.duplicate_nodes, filename=duplicate_nodes_path)
+    write_df(df=qc.dangling_edges, filename=dangling_edges_path)
diff --git a/cat_merge/merge.py b/cat_merge/merge.py
@@ -82,6 +82,8 @@ def merge(
         output_dir=output_dir
     )
 
+    write_qc(name=name, qc=qc, output_dir=output_dir)
+
     if qc_report:
         print("Generating QC report")
         qc_report = create_qc_report(kg, qc)