Summary Failure Message

 - Finds similar looking alleles
 - Summary message in batch reduce prints a table

Author: pbashyal-nmdp

pyard/db.py

@@ -138,6 +138,7 @@ def is_valid_mac_code(connection: sqlite3.Connection, code: str) -> bool:
     cursor.close()
     return result[0] > 0
 
+
 def serology_to_alleles(connection: sqlite3.Connection, serology: str) -> List[str]:
     """
     Look up Serology in the database and return corresponding list of alleles.
@@ -156,6 +157,7 @@ def serology_to_alleles(connection: sqlite3.Connection, serology: str) -> List[s
         alleles = []
     return alleles
 
+
 def is_valid_serology(connection: sqlite3.Connection, serology: str) -> bool:
     """
     Check db if the serology exists
@@ -287,3 +289,21 @@ def load_dict(connection: sqlite3.Connection, table_name: str, columns: Tuple[st
     table_as_dict = {k: v for k, v in cursor.fetchall()}
     cursor.close()
     return table_as_dict
+
+
+def similar_alleles(connection: sqlite3.Connection, allele_name: str) -> Set[str]:
+    """
+    Find similar alleles starting with the provided allele_name.
+
+    :param connection: db connection of type sqlite.Connection
+    :param allele_name: Allele name to use as a prefix to find similar alleles
+    :return: list of similar alleles
+    """
+    query = f"SELECT allele FROM alleles WHERE allele LIKE ?"
+    cursor = connection.execute(query, (f"{allele_name}%",))
+    result = cursor.fetchall()
+    # fetchall() returns a list of tuples of results
+    # e.g. [('C*04:09N',)]
+    # Get out the first value of the tuple from the result list
+    alleles = set(map(lambda t: t[0], result))
+    return alleles

scripts/pyard-reduce-csv

@@ -125,7 +125,7 @@ def reduce(allele, locus, column_name):
                 print(e)
             message = f"Failed reducing '{locus_allele}' in column {column_name}"
             print(message)
-            failure_summary_messages.append(message)
+            failed_to_reduce_alleles.append((column_name, locus_allele))
             return allele
         # print(f"reduced to '{reduced_allele}'")
         if reduced_allele:
@@ -210,7 +210,7 @@ if __name__ == '__main__':
                      header=0, dtype=str,
                      keep_default_na=False)
 
-    failure_summary_messages = []
+    failed_to_reduce_alleles = []
     # Reduce each of the specified columns
     for column in ard_config["columns_to_reduce_in_csv"]:
         if verbose:
@@ -245,12 +245,20 @@ if __name__ == '__main__':
         out_file_name = f"{ard_config['out_csv_filename'] + '.gz' if ard_config['apply_compression'] else ''}"
         df.to_csv(out_file_name, index=False, compression=ard_config["apply_compression"])
 
-    if len(failure_summary_messages) == 0:
+    if len(failed_to_reduce_alleles) == 0:
         print("No Errors", file=sys.stderr)
     else:
         print("Summary", file=sys.stderr)
         print("-------", file=sys.stderr)
-        for message in failure_summary_messages:
-            print("\t", message, file=sys.stderr)
+        print(f"{len(failed_to_reduce_alleles)} alleles failed to reduce.", file=sys.stderr)
+        print("| Column  Name    |      Allele      |      Did you mean ?       ", file=sys.stderr)
+        print("| --------------- | ---------------- | ------------------------- ", file=sys.stderr)
+        for column_name, locus_allele in failed_to_reduce_alleles:
+            similar_allele_names = pyard.db.similar_alleles(ard.db_connection, locus_allele)
+            if similar_allele_names:
+                similar_allele_names = ",".join(sorted(similar_allele_names, reverse=True))
+            else:
+                similar_allele_names = 'NA'
+            print(f"| {column_name:15} | {locus_allele:16} | {similar_allele_names} ", file=sys.stderr)
     # Done
     print(f"Saved result to file:{out_file_name}")