Update pandas syntax to avoid deprecated behavior (#3040)

seaborn/categorical.py

@@ -289,7 +289,7 @@ def plot_strips(
             jitter_move = jitterer(size=len(sub_data)) if len(sub_data) > 1 else 0
 
             adjusted_data = sub_data[self.cat_axis] + dodge_move + jitter_move
-            sub_data.loc[:, self.cat_axis] = adjusted_data
+            sub_data[self.cat_axis] = adjusted_data
 
             for var in "xy":
                 if self._log_scaled(var):
@@ -346,7 +346,7 @@ def plot_swarms(
                 dodge_move = offsets[sub_data["hue"].map(self._hue_map.levels.index)]
 
             if not sub_data.empty:
-                sub_data.loc[:, self.cat_axis] = sub_data[self.cat_axis] + dodge_move
+                sub_data[self.cat_axis] = sub_data[self.cat_axis] + dodge_move
 
             for var in "xy":
                 if self._log_scaled(var):

seaborn/distributions.py

@@ -246,30 +246,26 @@ def _resolve_multiple(self, curves, multiple):
 
             # Find column groups that are nested within col/row variables
             column_groups = {}
-            for i, keyd in enumerate(map(dict, curves.columns.tolist())):
+            for i, keyd in enumerate(map(dict, curves.columns)):
                 facet_key = keyd.get("col", None), keyd.get("row", None)
                 column_groups.setdefault(facet_key, [])
                 column_groups[facet_key].append(i)
 
             baselines = curves.copy()
-            for cols in column_groups.values():
+            for col_idxs in column_groups.values():
+                cols = curves.columns[col_idxs]
 
-                norm_constant = curves.iloc[:, cols].sum(axis="columns")
+                norm_constant = curves[cols].sum(axis="columns")
 
                 # Take the cumulative sum to stack
-                curves.iloc[:, cols] = curves.iloc[:, cols].cumsum(axis="columns")
+                curves[cols] = curves[cols].cumsum(axis="columns")
 
                 # Normalize by row sum to fill
                 if multiple == "fill":
-                    curves.iloc[:, cols] = (curves
-                                            .iloc[:, cols]
-                                            .div(norm_constant, axis="index"))
+                    curves[cols] = curves[cols].div(norm_constant, axis="index")
 
                 # Define where each segment starts
-                baselines.iloc[:, cols] = (curves
-                                           .iloc[:, cols]
-                                           .shift(1, axis=1)
-                                           .fillna(0))
+                baselines[cols] = curves[cols].shift(1, axis=1).fillna(0)
 
         if multiple == "dodge":
 

tests/_stats/test_histogram.py

@@ -157,7 +157,7 @@ def test_common_norm_subset(self, long_df, triple_args):
 
         h = Hist(stat="percent", common_norm=["a"])
         out = h(long_df, *triple_args)
-        for _, out_part in out.groupby(["a"]):
+        for _, out_part in out.groupby("a"):
             assert out_part["y"].sum() == pytest.approx(100)
 
     def test_common_bins_default(self, long_df, triple_args):
@@ -183,7 +183,7 @@ def test_common_bins_subset(self, long_df, triple_args):
         h = Hist(common_bins=False)
         out = h(long_df, *triple_args)
         bins = []
-        for _, out_part in out.groupby(["a"]):
+        for _, out_part in out.groupby("a"):
             bins.append(tuple(out_part["x"]))
         assert len(set(bins)) == out["a"].nunique()
 

tests/test_categorical.py

@@ -340,7 +340,7 @@ def test_longform_groupby(self):
         p1 = cat._CategoricalPlotter()
         p1.establish_variables(self.g, self.y, hue=self.h)
         p2 = cat._CategoricalPlotter()
-        p2.establish_variables(self.g, self.y[::-1], self.h)
+        p2.establish_variables(self.g, self.y.iloc[::-1], self.h)
         for i, (d1, d2) in enumerate(zip(p1.plot_data, p2.plot_data)):
             assert np.array_equal(d1.sort_index(), d2.sort_index())
 
@@ -614,7 +614,7 @@ def test_nested_stats(self):
                                       y.groupby([g, h]).mean().unstack())
 
         for ci_g, (_, grp_y) in zip(p.confint, y.groupby(g)):
-            for ci, hue_y in zip(ci_g, [grp_y[::2], grp_y[1::2]]):
+            for ci, hue_y in zip(ci_g, [grp_y.iloc[::2], grp_y.iloc[1::2]]):
                 sem = hue_y.std() / np.sqrt(len(hue_y))
                 mean = hue_y.mean()
                 half_ci = _normal_quantile_func(.975) * sem
@@ -732,7 +732,7 @@ def test_nested_sd_error_bars(self):
                                       y.groupby([g, h]).mean().unstack())
 
         for ci_g, (_, grp_y) in zip(p.confint, y.groupby(g)):
-            for ci, hue_y in zip(ci_g, [grp_y[::2], grp_y[1::2]]):
+            for ci, hue_y in zip(ci_g, [grp_y.iloc[::2], grp_y.iloc[1::2]]):
                 mean = hue_y.mean()
                 half_ci = np.std(hue_y)
                 ci_want = mean - half_ci, mean + half_ci