cleaner order for stacked hist

src/mplhep/plot.py

@@ -379,20 +379,6 @@ def iterable_not_string(arg):
         _chunked_kwargs = [_chunked_kwargs[ix] for ix in order]
         _labels = [_labels[ix] for ix in order]
 
-    elif stack:
-        # Sort from top to bottom so ax.legend() works as expected
-        order = np.argsort(label)[::-1]
-        plottables = [plottables[ix] for ix in order]
-        _chunked_kwargs = [_chunked_kwargs[ix] for ix in order]
-        _labels = [_labels[ix] for ix in order]
-        if "color" not in kwargs:
-            # Inverse default color cycle
-            _colors = itertools.cycle(
-                plt.rcParams["axes.prop_cycle"][len(plottables) - 1 :: -1]
-            )
-            for i in range(len(plottables)):
-                _chunked_kwargs[i].update(next(_colors))
-
     # ############################
     # # Stacking, norming, density
     if density is True and binwnorm is not None:
@@ -424,6 +410,19 @@ def iterable_not_string(arg):
     ##########
     # Plotting
     return_artists: list[StairsArtists | ErrorBarArtists] = []
+
+    if stack:
+        plottables = plottables[::-1]
+        _chunked_kwargs = _chunked_kwargs[::-1]
+        _labels = _labels[::-1]
+        if "color" not in kwargs:
+            # Inverse default color cycle
+            _colors = itertools.cycle(
+                plt.rcParams["axes.prop_cycle"][len(plottables) - 1 :: -1]
+            )
+            for i in range(len(plottables)):
+                _chunked_kwargs[i].update(next(_colors))
+
     if histtype == "step":
         for i in range(len(plottables)):
             do_errors = yerr is not False and (

src/mplhep/utils.py

@@ -294,8 +294,6 @@ def to_errorbar(self):
 
 
 def stack(*plottables):
-    # Sort from top to bottom so ax.legend() works as expected
-    plottables = plottables[::-1]
     baseline = np.nan_to_num(copy.deepcopy(plottables[0].values), 0)
     for i in range(1, len(plottables)):
         _mask = np.isnan(plottables[i].values)
@@ -305,7 +303,7 @@ def stack(*plottables):
         baseline += np.nan_to_num(plottables[i].values, 0)
         plottables[i].values = np.nansum([plottables[i].values, _baseline], axis=0)
         plottables[i].values[_mask] = np.nan
-    return plottables[::-1]
+    return plottables
 
 
 def align_marker(