BUG: DataFrame respects dtype with masked recarray (#24874)

doc/source/whatsnew/v0.24.0.rst

@@ -1692,8 +1692,8 @@ Missing
 - Bug in :func:`Series.hasnans` that could be incorrectly cached and return incorrect answers if null elements are introduced after an initial call (:issue:`19700`)
 - :func:`Series.isin` now treats all NaN-floats as equal also for ``np.object``-dtype. This behavior is consistent with the behavior for float64 (:issue:`22119`)
 - :func:`unique` no longer mangles NaN-floats and the ``NaT``-object for ``np.object``-dtype, i.e. ``NaT`` is no longer coerced to a NaN-value and is treated as a different entity. (:issue:`22295`)
-- :func:`DataFrame` and :func:`Series` now properly handle numpy masked arrays with hardened masks. Previously, constructing a DataFrame or Series from a masked array with a hard mask would create a pandas object containing the underlying value, rather than the expected NaN. (:issue:`24574`)
-
+- :class:`DataFrame` and :class:`Series` now properly handle numpy masked arrays with hardened masks. Previously, constructing a DataFrame or Series from a masked array with a hard mask would create a pandas object containing the underlying value, rather than the expected NaN. (:issue:`24574`)
+- Bug in :class:`DataFrame` constructor where ``dtype`` argument was not honored when handling numpy masked record arrays. (:issue:`24874`)
 
 MultiIndex
 ^^^^^^^^^^

pandas/core/internals/construction.py

@@ -93,7 +93,7 @@ def masked_rec_array_to_mgr(data, index, columns, dtype, copy):
     if columns is None:
         columns = arr_columns
 
-    mgr = arrays_to_mgr(arrays, arr_columns, index, columns)
+    mgr = arrays_to_mgr(arrays, arr_columns, index, columns, dtype)
 
     if copy:
         mgr = mgr.copy()

pandas/tests/frame/test_constructors.py

@@ -787,6 +787,17 @@ def test_constructor_maskedarray_hardened(self):
             dtype=float)
         tm.assert_frame_equal(result, expected)
 
+    def test_constructor_maskedrecarray_dtype(self):
+        # Ensure constructor honors dtype
+        data = np.ma.array(
+            np.ma.zeros(5, dtype=[('date', '<f8'), ('price', '<f8')]),
+            mask=[False] * 5)
+        data = data.view(ma.mrecords.mrecarray)
+        result = pd.DataFrame(data, dtype=int)
+        expected = pd.DataFrame(np.zeros((5, 2), dtype=int),
+                                columns=['date', 'price'])
+        tm.assert_frame_equal(result, expected)
+
     def test_constructor_mrecarray(self):
         # Ensure mrecarray produces frame identical to dict of masked arrays
         # from GH3479