BUG: Index.where casting ints to str (#37591)

Author: jbrockmendel

doc/source/whatsnew/v1.2.0.rst

@@ -456,6 +456,7 @@ Indexing
 - Bug in :meth:`Series.loc.__getitem__` with a non-unique :class:`MultiIndex` and an empty-list indexer (:issue:`13691`)
 - Bug in indexing on a :class:`Series` or :class:`DataFrame` with a :class:`MultiIndex` with a level named "0" (:issue:`37194`)
 - Bug in :meth:`Series.__getitem__` when using an unsigned integer array as an indexer giving incorrect results or segfaulting instead of raising ``KeyError`` (:issue:`37218`)
+- Bug in :meth:`Index.where` incorrectly casting numeric values to strings (:issue:`37591`)
 
 Missing
 ^^^^^^^

pandas/core/indexes/base.py

@@ -40,7 +40,6 @@
     ensure_int64,
     ensure_object,
     ensure_platform_int,
-    is_bool,
     is_bool_dtype,
     is_categorical_dtype,
     is_datetime64_any_dtype,
@@ -4079,23 +4078,16 @@ def where(self, cond, other=None):
         if other is None:
             other = self._na_value
 
-        dtype = self.dtype
         values = self.values
 
-        if is_bool(other) or is_bool_dtype(other):
-
-            # bools force casting
-            values = values.astype(object)
-            dtype = None
+        try:
+            self._validate_fill_value(other)
+        except (ValueError, TypeError):
+            return self.astype(object).where(cond, other)
 
         values = np.where(cond, values, other)
 
-        if self._is_numeric_dtype and np.any(isna(values)):
-            # We can't coerce to the numeric dtype of "self" (unless
-            # it's float) if there are NaN values in our output.
-            dtype = None
-
-        return Index(values, dtype=dtype, name=self.name)
+        return Index(values, name=self.name)
 
     # construction helpers
     @final

pandas/core/indexes/datetimelike.py

@@ -482,16 +482,9 @@ def isin(self, values, level=None):
 
     @Appender(Index.where.__doc__)
     def where(self, cond, other=None):
-        values = self._data._ndarray
+        other = self._data._validate_setitem_value(other)
 
-        try:
-            other = self._data._validate_setitem_value(other)
-        except (TypeError, ValueError) as err:
-            # Includes tzawareness mismatch and IncompatibleFrequencyError
-            oth = getattr(other, "dtype", other)
-            raise TypeError(f"Where requires matching dtype, not {oth}") from err
-
-        result = np.where(cond, values, other)
+        result = np.where(cond, self._data._ndarray, other)
         arr = self._data._from_backing_data(result)
         return type(self)._simple_new(arr, name=self.name)
 

pandas/core/indexes/numeric.py

@@ -121,6 +121,8 @@ def _validate_fill_value(self, value):
             # force conversion to object
             # so we don't lose the bools
             raise TypeError
+        if isinstance(value, str):
+            raise TypeError
 
         return value
 

pandas/tests/indexes/base_class/test_where.py

@@ -0,0 +1,13 @@
+import numpy as np
+
+from pandas import Index
+import pandas._testing as tm
+
+
+class TestWhere:
+    def test_where_intlike_str_doesnt_cast_ints(self):
+        idx = Index(range(3))
+        mask = np.array([True, False, True])
+        res = idx.where(mask, "2")
+        expected = Index([0, "2", 2])
+        tm.assert_index_equal(res, expected)

pandas/tests/indexes/datetimelike.py

@@ -143,10 +143,9 @@ def test_where_cast_str(self):
         result = index.where(mask, [str(index[0])])
         tm.assert_index_equal(result, expected)
 
-        msg = "Where requires matching dtype, not foo"
+        msg = "value should be a '.*', 'NaT', or array of those"
         with pytest.raises(TypeError, match=msg):
             index.where(mask, "foo")
 
-        msg = r"Where requires matching dtype, not \['foo'\]"
         with pytest.raises(TypeError, match=msg):
             index.where(mask, ["foo"])

pandas/tests/indexes/datetimes/test_indexing.py

@@ -177,24 +177,26 @@ def test_where_invalid_dtypes(self):
 
         i2 = Index([pd.NaT, pd.NaT] + dti[2:].tolist())
 
-        with pytest.raises(TypeError, match="Where requires matching dtype"):
+        msg = "value should be a 'Timestamp', 'NaT', or array of those. Got"
+        msg2 = "Cannot compare tz-naive and tz-aware datetime-like objects"
+        with pytest.raises(TypeError, match=msg2):
             # passing tz-naive ndarray to tzaware DTI
             dti.where(notna(i2), i2.values)
 
-        with pytest.raises(TypeError, match="Where requires matching dtype"):
+        with pytest.raises(TypeError, match=msg2):
             # passing tz-aware DTI to tznaive DTI
             dti.tz_localize(None).where(notna(i2), i2)
 
-        with pytest.raises(TypeError, match="Where requires matching dtype"):
+        with pytest.raises(TypeError, match=msg):
             dti.where(notna(i2), i2.tz_localize(None).to_period("D"))
 
-        with pytest.raises(TypeError, match="Where requires matching dtype"):
+        with pytest.raises(TypeError, match=msg):
             dti.where(notna(i2), i2.asi8.view("timedelta64[ns]"))
 
-        with pytest.raises(TypeError, match="Where requires matching dtype"):
+        with pytest.raises(TypeError, match=msg):
             dti.where(notna(i2), i2.asi8)
 
-        with pytest.raises(TypeError, match="Where requires matching dtype"):
+        with pytest.raises(TypeError, match=msg):
             # non-matching scalar
             dti.where(notna(i2), pd.Timedelta(days=4))
 
@@ -203,7 +205,7 @@ def test_where_mismatched_nat(self, tz_aware_fixture):
         dti = pd.date_range("2013-01-01", periods=3, tz=tz)
         cond = np.array([True, False, True])
 
-        msg = "Where requires matching dtype"
+        msg = "value should be a 'Timestamp', 'NaT', or array of those. Got"
         with pytest.raises(TypeError, match=msg):
             # wrong-dtyped NaT
             dti.where(cond, np.timedelta64("NaT", "ns"))

pandas/tests/indexes/period/test_indexing.py

@@ -545,24 +545,25 @@ def test_where_invalid_dtypes(self):
 
         i2 = PeriodIndex([NaT, NaT] + pi[2:].tolist(), freq="D")
 
-        with pytest.raises(TypeError, match="Where requires matching dtype"):
+        msg = "value should be a 'Period', 'NaT', or array of those"
+        with pytest.raises(TypeError, match=msg):
             pi.where(notna(i2), i2.asi8)
 
-        with pytest.raises(TypeError, match="Where requires matching dtype"):
+        with pytest.raises(TypeError, match=msg):
             pi.where(notna(i2), i2.asi8.view("timedelta64[ns]"))
 
-        with pytest.raises(TypeError, match="Where requires matching dtype"):
+        with pytest.raises(TypeError, match=msg):
             pi.where(notna(i2), i2.to_timestamp("S"))
 
-        with pytest.raises(TypeError, match="Where requires matching dtype"):
+        with pytest.raises(TypeError, match=msg):
             # non-matching scalar
             pi.where(notna(i2), Timedelta(days=4))
 
     def test_where_mismatched_nat(self):
         pi = period_range("20130101", periods=5, freq="D")
         cond = np.array([True, False, True, True, False])
 
-        msg = "Where requires matching dtype"
+        msg = "value should be a 'Period', 'NaT', or array of those"
         with pytest.raises(TypeError, match=msg):
             # wrong-dtyped NaT
             pi.where(cond, np.timedelta64("NaT", "ns"))

pandas/tests/indexes/timedeltas/test_indexing.py

@@ -150,24 +150,25 @@ def test_where_invalid_dtypes(self):
 
         i2 = Index([pd.NaT, pd.NaT] + tdi[2:].tolist())
 
-        with pytest.raises(TypeError, match="Where requires matching dtype"):
+        msg = "value should be a 'Timedelta', 'NaT', or array of those"
+        with pytest.raises(TypeError, match=msg):
             tdi.where(notna(i2), i2.asi8)
 
-        with pytest.raises(TypeError, match="Where requires matching dtype"):
+        with pytest.raises(TypeError, match=msg):
             tdi.where(notna(i2), i2 + pd.Timestamp.now())
 
-        with pytest.raises(TypeError, match="Where requires matching dtype"):
+        with pytest.raises(TypeError, match=msg):
             tdi.where(notna(i2), (i2 + pd.Timestamp.now()).to_period("D"))
 
-        with pytest.raises(TypeError, match="Where requires matching dtype"):
+        with pytest.raises(TypeError, match=msg):
             # non-matching scalar
             tdi.where(notna(i2), pd.Timestamp.now())
 
     def test_where_mismatched_nat(self):
         tdi = timedelta_range("1 day", periods=3, freq="D", name="idx")
         cond = np.array([True, False, False])
 
-        msg = "Where requires matching dtype"
+        msg = "value should be a 'Timedelta', 'NaT', or array of those"
         with pytest.raises(TypeError, match=msg):
             # wrong-dtyped NaT
             tdi.where(cond, np.datetime64("NaT", "ns"))

pandas/tests/indexing/test_coercion.py

@@ -780,7 +780,7 @@ def test_where_index_timedelta64(self, value):
         result = tdi.where(cond, value)
         tm.assert_index_equal(result, expected)
 
-        msg = "Where requires matching dtype"
+        msg = "value should be a 'Timedelta', 'NaT', or array of thos"
         with pytest.raises(TypeError, match=msg):
             # wrong-dtyped NaT
             tdi.where(cond, np.datetime64("NaT", "ns"))
@@ -804,11 +804,12 @@ def test_where_index_period(self):
         tm.assert_index_equal(result, expected)
 
         # Passing a mismatched scalar
-        msg = "Where requires matching dtype"
+        msg = "value should be a 'Period', 'NaT', or array of those"
         with pytest.raises(TypeError, match=msg):
             pi.where(cond, pd.Timedelta(days=4))
 
-        with pytest.raises(TypeError, match=msg):
+        msg = r"Input has different freq=D from PeriodArray\(freq=Q-DEC\)"
+        with pytest.raises(ValueError, match=msg):
             pi.where(cond, pd.Period("2020-04-21", "D"))