ENH set default colormap

sharp/visualization/_aggregate.py

@@ -15,7 +15,7 @@ def strata_boxplots(
     feature_names=None,
     n_strata=5,
     gap_size=1,
-    cmap=None,
+    cmap="Pastel1",
     ax=None,
     **kwargs,
 ):
@@ -50,38 +50,37 @@ def strata_boxplots(
 
     df["target_binned"] = df["target_binned"].str.replace("<", "$<$")
 
-    colors = [
-        plt.get_cmap(cmap)(i / len(feature_names)) for i in range(len(feature_names))
-    ]
+    colors = [plt.get_cmap(cmap)(i) for i in range(len(feature_names))]
     bin_names = df["target_binned"].unique()
     pos_increment = 1 / (len(feature_names) + gap_size)
     boxes = []
-
     for i, bin_name in enumerate(bin_names):
-        box = ax.boxplot(
-            [df[df["target_binned"] == bin_name][feature] for feature in feature_names],
+        box = plt.boxplot(
+            df[df["target_binned"] == bin_name][feature_names],
             widths=pos_increment,
             positions=[i + pos_increment * n for n in range(len(feature_names))],
             patch_artist=True,
             medianprops={"color": "black"},
+            boxprops={"facecolor": "C0", "edgecolor": "black"},
             **kwargs,
         )
         boxes.append(box)
 
-    for feature_idx, color in enumerate(colors):
-        for i, box in enumerate(boxes):
-            box["boxes"][feature_idx].set_facecolor(color)
+    for box in boxes:
+        patches = []
+        for patch, color in zip(box["boxes"], colors):
+            patch.set_facecolor(color)
+            patches.append(patch)
 
-    ax.set_xticks(
-        np.arange(0, len(bin_names)) + pos_increment * (len(feature_names) - 1) / 2
+    plt.xticks(
+        np.arange(0, len(bin_names)) + pos_increment * (len(feature_names) - 1) / 2,
+        bin_names,
     )
-    ax.set_xticklabels(bin_names)
 
-    patches = [plt.Line2D([0], [0], color=color, lw=4) for color in colors]
-    ax.legend(
+    plt.legend(
         patches,
         feature_names,
-        loc="upper center",
+        loc="lower center",
         bbox_to_anchor=(0.5, 1.05),
         ncol=len(feature_names),
     )

sharp/visualization/_visualization.py

@@ -10,7 +10,7 @@
 from ._aggregate import strata_boxplots
 
 
-class ShaRPViz:  # TODO
+class ShaRPViz:
     def __init__(self, sharp):
         self.sharp = sharp
 
@@ -54,7 +54,7 @@ def strata_boxplot(
         feature_names=None,
         n_strata=5,
         gap_size=1,
-        cmap=None,
+        cmap="Pastel1",
         ax=None,
         **kwargs
     ):