Replaced deprecated _construct_axes_from_arguments method in pandas 2…

….x.x and fixed quantile keyword argument Signed-off-by: Yerzhaisang <tasqali1697@gmail.com>
opensearch-project · Feb 13, 2024 · 4067fcf · 4067fcf
1 parent 24163a0
commit 4067fcf
Show file tree

Hide file tree

Showing 3 changed files with 18 additions and 10 deletions.
diff --git a/opensearch_py_ml/dataframe.py b/opensearch_py_ml/dataframe.py
@@ -424,9 +424,7 @@ def drop(
             axis = pd.DataFrame._get_axis_name(axis)
             axes = {axis: labels}
         elif index is not None or columns is not None:
-            axes, _ = pd.DataFrame()._construct_axes_from_arguments(
-                (index, columns), {}
-            )
+            axes = {'index': index, 'columns': columns}
         else:
             raise ValueError(
                 "Need to specify at least one of 'labels', 'index' or 'columns'"

diff --git a/tests/common.py b/tests/common.py
@@ -164,3 +164,6 @@ def mad(x):
         numeric_columns = x.select_dtypes(include=["number", "bool"]).columns
         x = x[numeric_columns]
     return np.fabs(x - x.mean()).mean()
+
+def quantile(x, numeric_only=None):
+    return x.quantile()
diff --git a/tests/dataframe/test_metrics_pytest.py b/tests/dataframe/test_metrics_pytest.py
@@ -29,7 +29,7 @@
 import pytest
 from pandas.testing import assert_frame_equal, assert_series_equal
 
-from tests.common import TestData, assert_almost_equal
+from tests.common import TestData, assert_almost_equal, quantile, mad
 
 
 class TestDataFrameMetrics(TestData):
@@ -81,9 +81,10 @@ def test_flights_extended_metrics(self):
         logger.setLevel(logging.DEBUG)
 
         for func in self.extended_funcs:
-            pd_metric = getattr(pd_flights, func)(
-                **({"numeric_only": True} if func != "mad" else {})
-            )
+            if func == "mad":
+                pd_metric = mad(pd_flights)
+            else:
+                pd_metric = getattr(pd_flights, func)(numeric_only=True)
             oml_metric = getattr(oml_flights, func)(numeric_only=True)
 
             pd_value = pd_metric["AvgTicketPrice"]
@@ -101,7 +102,10 @@ def test_flights_extended_metrics_nan(self):
         ]
 
         for func in self.extended_funcs:
-            pd_metric = getattr(pd_flights_1, func)()
+            if func == "mad":
+                pd_metric = mad(pd_flights_1)
+            else:
+                pd_metric = getattr(pd_flights_1, func)()
             oml_metric = getattr(oml_flights_1, func)(numeric_only=False)
 
             assert_series_equal(pd_metric, oml_metric, check_exact=False)
@@ -111,7 +115,10 @@ def test_flights_extended_metrics_nan(self):
         oml_flights_0 = oml_flights[oml_flights.FlightNum == "XXX"][["AvgTicketPrice"]]
 
         for func in self.extended_funcs:
-            pd_metric = getattr(pd_flights_0, func)()
+            if func == "mad":
+                pd_metric = mad(pd_flights_0)
+            else:
+                pd_metric = getattr(pd_flights_0, func)()
             oml_metric = getattr(oml_flights_0, func)(numeric_only=False)
 
             assert_series_equal(pd_metric, oml_metric, check_exact=False)
@@ -498,7 +505,7 @@ def test_flights_agg_quantile(self, numeric_only):
             ["AvgTicketPrice", "FlightDelayMin", "dayOfWeek"]
         )
 
-        pd_quantile = pd_flights.agg(["quantile", "min"], numeric_only=numeric_only)
+        pd_quantile = pd_flights.agg([quantile, "min"], numeric_only=numeric_only)
         oml_quantile = oml_flights.agg(["quantile", "min"], numeric_only=numeric_only)
 
         assert_frame_equal(