Adds violinplot.

Author: Niru Maheswaranathan

jetplot/__init__.py

@@ -1,6 +1,6 @@
 """Jetpack is a set of useful utility functions for scientific python."""
 
-__version__ = "0.6.0"
+__version__ = "0.6.2"
 
 from . import colors as c, typing
 from .chart_utils import *

jetplot/images.py

@@ -134,7 +134,7 @@ def cmat(
         ax.set_xticks(np.arange(num_cols))
         ax.set_xticklabels(labels, rotation=90, fontsize=label_fontsize)
         ax.set_yticks(np.arange(num_rows))
-        ax.set_yticklabels(labels, label_fontsize=label_fontsize)
+        ax.set_yticklabels(labels, fontsize=label_fontsize)
 
     ax.xaxis.set_minor_locator(FixedLocator(np.arange(num_cols) - 0.5))
     ax.yaxis.set_minor_locator(FixedLocator(np.arange(num_rows) - 0.5))

jetplot/plots.py

@@ -1,20 +1,20 @@
 """Common plots."""
 
 import numpy as np
-from scipy.stats import gaussian_kde
-from sklearn.covariance import EmpiricalCovariance, MinCovDet
-
 from matplotlib.patches import Ellipse
 from matplotlib.transforms import Affine2D
+from scipy.stats import gaussian_kde
+from sklearn.covariance import EmpiricalCovariance, MinCovDet
 
 from .chart_utils import figwrapper, nospines, plotwrapper
-from .colors import cmap_colors
+from .colors import cmap_colors, neutral
 from .typing import Color
 
 __all__ = [
     "hist",
     "hist2d",
     "errorplot",
+    "violinplot",
     "bar",
     "lines",
     "waterfall",
@@ -23,6 +23,67 @@
 ]
 
 
+@plotwrapper
+def violinplot(
+    data,
+    xs,
+    fc=neutral[3],
+    ec=neutral[9],
+    mc=neutral[1],
+    showmedians=True,
+    showmeans=False,
+    showquartiles=True,
+    **kwargs,
+):
+    _ = kwargs.pop("fig")
+    ax = kwargs.pop("ax")
+
+    data = np.atleast_2d(data).T
+
+    if isinstance(xs, float) or isinstance(xs, int):
+        xs = [
+            xs,
+        ]
+
+    parts = ax.violinplot(
+        data, positions=xs, showmeans=False, showmedians=False, showextrema=False
+    )
+
+    for pc in parts["bodies"]:
+        pc.set_facecolor(fc)
+        pc.set_edgecolor(ec)
+        pc.set_alpha(1.0)
+
+    q1, medians, q3 = np.percentile(data, [25, 50, 75], axis=0)
+
+    ax.vlines(
+        xs,
+        np.min(data, axis=0),
+        np.max(data, axis=0),
+        color=ec,
+        linestyle="-",
+        lw=1,
+        zorder=10,
+        label="Extrema",
+    )
+
+    if showquartiles:
+        ax.vlines(xs, q1, q3, color=ec, linestyle="-", lw=5, zorder=5)
+
+    if showmedians:
+        ax.scatter(xs, medians, marker="o", color=mc, s=15, zorder=20)
+
+    if showmeans:
+        ax.scatter(
+            xs,
+            np.mean(data, axis=0),
+            marker="s",
+            color=mc,
+            s=15,
+            zorder=20,
+        )
+
+
 @plotwrapper
 def hist(*args, **kwargs):
     """Wrapper for matplotlib.hist function."""
@@ -81,7 +142,7 @@ def errorplot(
     err_color: Color = "#cccccc",
     alpha_fill=1.0,
     clip_on=True,
-    **kwargs
+    **kwargs,
 ):
     """Plot a line with error bars."""
     ax = kwargs["ax"]
@@ -129,7 +190,15 @@ def errorplot(
 
 @plotwrapper
 def bar(
-    labels, data, color="#888888", width=0.7, offset=0.0, err=None, capsize=5, capthick=2, **kwargs
+    labels,
+    data,
+    color="#888888",
+    width=0.7,
+    offset=0.0,
+    err=None,
+    capsize=5,
+    capthick=2,
+    **kwargs,
 ):
     """Bar chart.
 
@@ -279,7 +348,7 @@ def ellipse(x, y, n_std=3.0, facecolor="none", estimator="empirical", **kwargs):
         width=ell_radius_x * 2,
         height=ell_radius_y * 2,
         facecolor=facecolor,
-        **kwargs
+        **kwargs,
     )
 
     # Calculating the stdandard deviation of x from