feat: add 'columns' as an alias for 'col_order'

docs/reading.rst

@@ -28,7 +28,8 @@ destination DataFrame as well as a preferred column order as follows:
        'SELECT * FROM `test_dataset.test_table`',
        project_id=projectid,
        index_col='index_column_name',
-       col_order=['col1', 'col2', 'col3'])
+       col_order=['col1', 'col2', 'col3'],
+       columns=['col1', 'col2'])
 
 Querying with legacy SQL syntax
 -------------------------------

pandas_gbq/gbq.py

@@ -733,6 +733,7 @@ def read_gbq(
     project_id=None,
     index_col=None,
     col_order=None,
+    columns=None,
     reauth=False,
     auth_local_webserver=True,
     dialect=None,
@@ -774,6 +775,8 @@ def read_gbq(
     col_order : list(str), optional
         List of BigQuery column names in the desired order for results
         DataFrame.
+    columns : list(str), optional
+        List of BigQuery column names to return, alias for col_order
     reauth : boolean, default False
         Force Google BigQuery to re-authenticate the user. This is useful
         if multiple accounts are used.
@@ -964,6 +967,9 @@ def read_gbq(
                 'Index column "{0}" does not exist in DataFrame.'.format(index_col)
             )
 
+    # Creating an alias for col_order, which is columns
+    col_order = col_order or columns
+
     # Change the order of columns in the DataFrame based on provided list
     if col_order is not None:
         if sorted(col_order) == sorted(final_df.columns):

tests/system/test_gbq.py

@@ -600,6 +600,49 @@ def test_tokyo(self, tokyo_dataset, tokyo_table, project_id):
         )
         assert df["max_year"][0] >= 2000
 
+    def test_columns_and_col_order(self, project_id):
+        query = "SELECT 'a' AS string_1, 'b' AS string_2, 'c' AS string_3"
+        columns = ["string_2", "string_1"]
+        col_order = ["string_3", "string_1", "string_2"]
+        result_frame = gbq.read_gbq(
+            query,
+            project_id=project_id,
+            columns=columns,
+            col_order=col_order,
+            credentials=self.credentials,
+            dialect="standard",
+        )
+        correct_frame = DataFrame(
+            {"string_1": ["a"], "string_2": ["b"], "string_3": ["c"]}
+        )[col_order]
+        tm.assert_frame_equal(result_frame, correct_frame)
+
+        # Verify that col_order is prioritized over columns
+        assert sorted(col_order) == sorted(result_frame.columns)
+
+    def test_col_order_as_alias(self, project_id):
+        query = "SELECT 'a' AS string_1, 'b' AS string_2, 'c' AS string_3"
+        columns = ["string_2", "string_1", "string_3"]
+        # Not explicitly specifying col_order
+
+        result_frame = gbq.read_gbq(
+            query,
+            project_id=project_id,
+            columns=columns,
+            credentials=self.credentials,
+            dialect="standard",
+        )
+
+        correct_frame = DataFrame(
+            {"string_1": ["a"], "string_2": ["b"], "string_3": ["c"]}
+        )[columns]
+
+        # Verify that the result_frame matches the expected DataFrame
+        tm.assert_frame_equal(result_frame, correct_frame)
+
+        # Ensure that the order of columns in the result_frame matches the specified order
+        assert sorted(columns) == sorted(result_frame.columns)
+
 
 class TestToGBQIntegration(object):
     @pytest.fixture(autouse=True, scope="function")