Backport PR pandas-dev#57233 on branch 2.2.x (REGR: Fix to_numpy conversion for arrow ea with float dtype given) (pandas-dev#57294)

Backport PR pandas-dev#57233: REGR: Fix to_numpy conversion for arrow ea with float dtype given

Co-authored-by: Patrick Hoefler <[email protected]>

Author: meeseeksmachine

doc/source/whatsnew/v2.2.1.rst

@@ -25,6 +25,7 @@ Fixed regressions
 - Fixed regression in :meth:`DataFrameGroupBy.idxmin`, :meth:`DataFrameGroupBy.idxmax`, :meth:`SeriesGroupBy.idxmin`, :meth:`SeriesGroupBy.idxmax` where values containing the minimum or maximum value for the dtype could produce incorrect results (:issue:`57040`)
 - Fixed regression in :meth:`Index.join` raising ``TypeError`` when joining an empty index to a non-empty index containing mixed dtype values (:issue:`57048`)
 - Fixed regression in :meth:`Series.pct_change` raising a ``ValueError`` for an empty :class:`Series` (:issue:`57056`)
+- Fixed regression in :meth:`Series.to_numpy` when dtype is given as float and the data contains NaNs (:issue:`57121`)
 
 .. ---------------------------------------------------------------------------
 .. _whatsnew_221.bug_fixes:

pandas/core/arrays/_utils.py

@@ -39,6 +39,8 @@ def to_numpy_dtype_inference(
             dtype = arr.dtype.numpy_dtype  # type: ignore[union-attr]
     elif dtype is not None:
         dtype = np.dtype(dtype)
+        if na_value is lib.no_default and hasna and dtype.kind == "f":
+            na_value = np.nan
         dtype_given = True
     else:
         dtype_given = True

pandas/tests/arrays/boolean/test_construction.py

@@ -308,8 +308,6 @@ def test_to_numpy(box):
     # converting to int or float without specifying na_value raises
     with pytest.raises(ValueError, match="cannot convert to 'int64'-dtype"):
         arr.to_numpy(dtype="int64")
-    with pytest.raises(ValueError, match="cannot convert to 'float64'-dtype"):
-        arr.to_numpy(dtype="float64")
 
 
 def test_to_numpy_copy():

pandas/tests/arrays/floating/test_to_numpy.py

@@ -33,10 +33,10 @@ def test_to_numpy_float(box):
     tm.assert_numpy_array_equal(result, expected)
 
     arr = con([0.1, 0.2, None], dtype="Float64")
-    with pytest.raises(ValueError, match="cannot convert to 'float64'-dtype"):
-        result = arr.to_numpy(dtype="float64")
+    result = arr.to_numpy(dtype="float64")
+    expected = np.array([0.1, 0.2, np.nan], dtype="float64")
+    tm.assert_numpy_array_equal(result, expected)
 
-    # need to explicitly specify na_value
     result = arr.to_numpy(dtype="float64", na_value=np.nan)
     expected = np.array([0.1, 0.2, np.nan], dtype="float64")
     tm.assert_numpy_array_equal(result, expected)
@@ -100,7 +100,7 @@ def test_to_numpy_dtype(box, dtype):
     tm.assert_numpy_array_equal(result, expected)
 
 
-@pytest.mark.parametrize("dtype", ["float64", "float32", "int32", "int64", "bool"])
+@pytest.mark.parametrize("dtype", ["int32", "int64", "bool"])
 @pytest.mark.parametrize("box", [True, False], ids=["series", "array"])
 def test_to_numpy_na_raises(box, dtype):
     con = pd.Series if box else pd.array

pandas/tests/arrays/integer/test_dtypes.py

@@ -271,7 +271,7 @@ def test_to_numpy_dtype(dtype, in_series):
     tm.assert_numpy_array_equal(result, expected)
 
 
-@pytest.mark.parametrize("dtype", ["float64", "int64", "bool"])
+@pytest.mark.parametrize("dtype", ["int64", "bool"])
 def test_to_numpy_na_raises(dtype):
     a = pd.array([0, 1, None], dtype="Int64")
     with pytest.raises(ValueError, match=dtype):

pandas/tests/series/methods/test_to_numpy.py

@@ -1,6 +1,8 @@
 import numpy as np
 import pytest
 
+import pandas.util._test_decorators as td
+
 from pandas import (
     NA,
     Series,
@@ -23,3 +25,12 @@ def test_to_numpy_cast_before_setting_na():
     result = ser.to_numpy(dtype=np.float64, na_value=np.nan)
     expected = np.array([1.0])
     tm.assert_numpy_array_equal(result, expected)
+
+
+@td.skip_if_no("pyarrow")
+def test_to_numpy_arrow_dtype_given():
+    # GH#57121
+    ser = Series([1, NA], dtype="int64[pyarrow]")
+    result = ser.to_numpy(dtype="float64")
+    expected = np.array([1.0, np.nan])
+    tm.assert_numpy_array_equal(result, expected)