extend to_gbq docstring and add tests for table_schema parameter

Author: mremes

pandas_gbq/gbq.py

@@ -816,6 +816,11 @@ def to_gbq(dataframe, destination_table, project_id, chunksize=10000,
         Service account private key in JSON format. Can be file path
         or string contents. This is useful for remote server
         authentication (eg. jupyter iPython notebook on remote host)
+    table_schema : list of dicts
+        List of BigQuery table fields to which according DataFrame columns
+        conform to, e.g. `[{'name': 'col1', 'type': 'STRING'},...]`. If
+        schema is not provided, it will be generated according to dtypes
+        of DataFrame columns.
     """
 
     if if_exists not in ('fail', 'replace', 'append'):

pandas_gbq/tests/test_gbq.py

@@ -1258,6 +1258,31 @@ def test_verify_schema_ignores_field_mode(self):
         assert self.sut.verify_schema(
             self.dataset_prefix + "1", TABLE_ID + test_id, test_schema_2)
 
+    def test_upload_data_with_valid_user_schema(self):
+        df = tm.makeMixedDataFrame()
+        test_id = "15"
+        test_schema = [{'name': 'A', 'type': 'FLOAT'},
+                       {'name': 'B', 'type': 'FLOAT'},
+                       {'name': 'C', 'type': 'STRING'},
+                       {'name': 'D', 'type': 'TIMESTAMP'}]
+        destination_table = self.destination_table + test_id
+        gbq.to_gbq(df, destination_table, _get_project_id(),
+                   private_key=_get_private_key_path(), table_schema=test_schema)
+        dataset, table = destination_table.split('.')
+        assert self.table.verify_schema(dataset, table, dict(fields=test_schema))
+
+    def test_upload_data_with_invalid_user_schema_raises_error(self):
+        df = tm.makeMixedDataFrame()
+        test_id = "16"
+        test_schema = [{'name': 'A', 'type': 'FLOAT'},
+                       {'name': 'B', 'type': 'FLOAT'},
+                       {'name': 'C', 'type': 'FLOAT'},
+                       {'name': 'D', 'type': 'FLOAT'}]
+        destination_table = self.destination_table + test_id
+        with tm.assertRaises(gbq.StreamingInsertError):
+            gbq.to_gbq(df, destination_table, _get_project_id(),
+                       private_key=_get_private_key_path(), table_schema=test_schema)
+
     def test_list_dataset(self):
         dataset_id = self.dataset_prefix + "1"
         assert dataset_id in self.dataset.datasets()