FIX-#2285: Default to pandas warning message improved

modin/backends/base/query_compiler.py

@@ -23,6 +23,7 @@
     CatDefault,
     GroupByDefault,
 )
+from modin.error_message import ErrorMessage
 
 from pandas.core.dtypes.common import is_scalar
 import pandas.core.resample
@@ -32,6 +33,7 @@
 
 def _get_axis(axis):
     def axis_getter(self):
+        ErrorMessage.default_to_pandas(f"DataFrame.get_axis({axis})")
         return self.to_pandas().axes[axis]
 
     return axis_getter

modin/backends/pandas/query_compiler.py

@@ -209,7 +209,8 @@ def default_to_pandas(self, pandas_op, *args, **kwargs):
         ----
         This operation takes a distributed object and converts it directly to pandas.
         """
-        ErrorMessage.default_to_pandas(str(pandas_op))
+        op_name = getattr(pandas_op, "__name__", str(pandas_op))
+        ErrorMessage.default_to_pandas(op_name)
         args = (a.to_pandas() if isinstance(a, type(self)) else a for a in args)
         kwargs = {
             k: v.to_pandas if isinstance(v, type(self)) else v

modin/data_management/functions/default_methods/binary_default.py

@@ -19,7 +19,7 @@
 
 class BinaryDefault(AnyDefault):
     @classmethod
-    def build_default_to_pandas(cls, fn):
+    def build_default_to_pandas(cls, fn, fn_name):
         def bin_ops_wrapper(df, other, *args, **kwargs):
             squeeze_other = kwargs.pop("broadcast", False) or kwargs.pop(
                 "squeeze_other", False
@@ -41,7 +41,4 @@ def bin_ops_wrapper(df, other, *args, **kwargs):
                 result = pandas.DataFrame(result)
             return result
 
-        def wrapper(self, *args, **kwargs):
-            return self.default_to_pandas(bin_ops_wrapper, *args, **kwargs)
-
-        return wrapper
+        return super().build_default_to_pandas(bin_ops_wrapper, fn_name)

modin/data_management/functions/default_methods/default.py

@@ -19,10 +19,13 @@
 
 
 class DefaultMethod(Function):
+    OBJECT_TYPE = "DataFrame"
+
     @classmethod
     def call(cls, func, **call_kwds):
         obj = call_kwds.get("obj_type", pandas.DataFrame)
         force_inplace = call_kwds.get("inplace")
+        fn_name = call_kwds.get("fn_name", getattr(func, "__name__", str(func)))
 
         if isinstance(func, str):
             fn = getattr(obj, func)
@@ -57,28 +60,21 @@ def applyier(df, *args, **kwargs):
                 inplace = force_inplace
             return result if not inplace else df
 
-        return cls.build_wrapper(applyier, func)
+        return cls.build_wrapper(applyier, fn_name)
 
     @classmethod
     def register(cls, func, **kwargs):
         return cls.call(func, **kwargs)
 
     @classmethod
-    def build_wrapper(cls, fn, fn_name=None):
-        wrapper = cls.build_default_to_pandas(fn)
+    def build_wrapper(cls, fn, fn_name):
+        wrapper = cls.build_default_to_pandas(fn, fn_name)
 
         def args_cast(self, *args, **kwargs):
             args = try_cast_to_pandas(args)
             kwargs = try_cast_to_pandas(kwargs)
             return wrapper(self, *args, **kwargs)
 
-        if fn_name is None:
-            fn_name = fn.__name__
-        if not isinstance(fn_name, str):
-            fn_name = getattr(fn_name, "__name__", repr(fn_name))
-
-        # setting proper function name that will be printed in default to pandas warning
-        args_cast.__name__ = fn_name
         return args_cast
 
     @classmethod
@@ -89,7 +85,9 @@ def property_wrapper(df):
         return property_wrapper
 
     @classmethod
-    def build_default_to_pandas(cls, fn):
+    def build_default_to_pandas(cls, fn, fn_name):
+        fn.__name__ = f"<function {cls.OBJECT_TYPE}.{fn_name}>"
+
         def wrapper(self, *args, **kwargs):
             return self.default_to_pandas(fn, *args, **kwargs)
 

modin/data_management/functions/default_methods/groupby_default.py

@@ -145,6 +145,8 @@ def build_groupby(cls, func):
 
 
 class GroupByDefault(DefaultMethod):
+    OBJECT_TYPE = "GroupBy"
+
     @classmethod
     def register(cls, func, **kwargs):
-        return cls.call(GroupBy.build_groupby(func), **kwargs)
+        return cls.call(GroupBy.build_groupby(func), fn_name=func.__name__, **kwargs)

modin/data_management/functions/default_methods/resample_default.py

@@ -31,6 +31,12 @@ def fn(df, resample_args, *args, **kwargs):
 
 
 class ResampleDefault(DefaultMethod):
+    OBJECT_TYPE = "Resampler"
+
     @classmethod
     def register(cls, func, squeeze_self=False, **kwargs):
-        return cls.call(Resampler.build_resample(func, squeeze_self), **kwargs)
+        return cls.call(
+            Resampler.build_resample(func, squeeze_self),
+            fn_name=func.__name__,
+            **kwargs
+        )

modin/data_management/functions/default_methods/rolling_default.py

@@ -29,6 +29,8 @@ def fn(df, rolling_args, *args, **kwargs):
 
 
 class RollingDefault(DefaultMethod):
+    OBJECT_TYPE = "Rolling"
+
     @classmethod
     def register(cls, func, **kwargs):
-        return cls.call(Rolling.build_rolling(func), **kwargs)
+        return cls.call(Rolling.build_rolling(func), fn_name=func.__name__, **kwargs)

modin/data_management/functions/default_methods/series_default.py

@@ -15,6 +15,8 @@
 
 
 class SeriesDefault(AnyDefault):
+    OBJECT_TYPE = "Series"
+
     @classmethod
     def frame_wrapper(cls, df):
         return df.squeeze(axis=1)

modin/pandas/test/test_general.py

@@ -16,6 +16,7 @@
 import modin.pandas as pd
 import numpy as np
 from numpy.testing import assert_array_equal
+from modin.utils import get_current_backend
 
 from .utils import test_data_values, test_data_keys, df_equals
 
@@ -437,3 +438,29 @@ def test_to_pandas_indices():
         assert md_df.axes[axis].equal_levels(
             pd_df.axes[axis]
         ), f"Levels of indices at axis {axis} are different!"
+
+
+@pytest.mark.skipif(
+    get_current_backend() != "BaseOnPython",
+    reason="This test make sense only on BaseOnPython backend.",
+)
+@pytest.mark.parametrize(
+    "func, regex",
+    [
+        (lambda df: df.mean(level=0), r"[^D]*DataFrame\.mean"),
+        (lambda df: df + df, r"[^D]*DataFrame\.add"),
+        (lambda df: df.index, r"[^D]*DataFrame.get_axis\(0\)"),
+        (
+            lambda df: df.drop(columns="col1").squeeze().repeat(2),
+            r"[^S]*Series\.repeat",
+        ),
+        (lambda df: df.groupby("col1").prod(), r"[^G]*GroupBy\.prod"),
+        (lambda df: df.rolling(1).count(), r"[^R]*Rolling\.count"),
+    ],
+)
+def test_default_to_pandas_warning_message(func, regex):
+    data = {"col1": [1, 2, 3], "col2": [4, 5, 6]}
+    df = pd.DataFrame(data)
+
+    with pytest.warns(UserWarning, match=regex):
+        func(df)