BUG: histogram weights aren't dropped if NaN values in data (#48888)

Author: AdamOrmondroyd

doc/source/whatsnew/v2.0.0.rst

@@ -926,6 +926,7 @@ Period
 
 Plotting
 ^^^^^^^^
+- Bug in :meth:`DataFrame.plot.hist`, not dropping elements of ``weights`` corresponding to ``NaN`` values in ``data`` (:issue:`48884`)
 - ``ax.set_xlim`` was sometimes raising ``UserWarning`` which users couldn't address due to ``set_xlim`` not accepting parsing arguments - the converter now uses :func:`Timestamp` instead (:issue:`49148`)
 -
 

pandas/plotting/_matplotlib/hist.py

@@ -146,14 +146,23 @@ def _make_plot(self) -> None:
                 kwds["label"] = self.columns
                 kwds.pop("color")
 
-            y = reformat_hist_y_given_by(y, self.by)
-
             # We allow weights to be a multi-dimensional array, e.g. a (10, 2) array,
             # and each sub-array (10,) will be called in each iteration. If users only
             # provide 1D array, we assume the same weights is used for all iterations
             weights = kwds.get("weights", None)
-            if weights is not None and np.ndim(weights) != 1:
-                kwds["weights"] = weights[:, i]
+            if weights is not None:
+                if np.ndim(weights) != 1 and np.shape(weights)[-1] != 1:
+                    try:
+                        weights = weights[:, i]
+                    except IndexError as err:
+                        raise ValueError(
+                            "weights must have the same shape as data, "
+                            "or be a single column"
+                        ) from err
+                weights = weights[~isna(y)]
+                kwds["weights"] = weights
+
+            y = reformat_hist_y_given_by(y, self.by)
 
             artists = self._plot(ax, y, column_num=i, stacking_id=stacking_id, **kwds)
 

pandas/tests/plotting/test_hist_method.py

@@ -560,6 +560,36 @@ def test_hist_secondary_legend(self):
         assert ax.get_yaxis().get_visible()
         tm.close()
 
+    @td.skip_if_no_mpl
+    def test_hist_with_nans_and_weights(self):
+        # GH 48884
+        df = DataFrame(
+            [[np.nan, 0.2, 0.3], [0.4, np.nan, np.nan], [0.7, 0.8, 0.9]],
+            columns=list("abc"),
+        )
+        weights = np.array([0.25, 0.3, 0.45])
+        no_nan_df = DataFrame([[0.4, 0.2, 0.3], [0.7, 0.8, 0.9]], columns=list("abc"))
+        no_nan_weights = np.array([[0.3, 0.25, 0.25], [0.45, 0.45, 0.45]])
+
+        from matplotlib.patches import Rectangle
+
+        _, ax0 = self.plt.subplots()
+        df.plot.hist(ax=ax0, weights=weights)
+        rects = [x for x in ax0.get_children() if isinstance(x, Rectangle)]
+        heights = [rect.get_height() for rect in rects]
+        _, ax1 = self.plt.subplots()
+        no_nan_df.plot.hist(ax=ax1, weights=no_nan_weights)
+        no_nan_rects = [x for x in ax1.get_children() if isinstance(x, Rectangle)]
+        no_nan_heights = [rect.get_height() for rect in no_nan_rects]
+        assert all(h0 == h1 for h0, h1 in zip(heights, no_nan_heights))
+
+        idxerror_weights = np.array([[0.3, 0.25], [0.45, 0.45]])
+
+        msg = "weights must have the same shape as data, or be a single column"
+        with pytest.raises(ValueError, match=msg):
+            _, ax2 = self.plt.subplots()
+            no_nan_df.plot.hist(ax=ax2, weights=idxerror_weights)
+
 
 @td.skip_if_no_mpl
 class TestDataFrameGroupByPlots(TestPlotBase):