wehub-resource-sync
82 lines
2.7 KiB
Python
82 lines
2.7 KiB
Python
import matplotlib.pyplot as plt
|
|
import numpy as np
|
|
import pandas as pd
|
|
import scipy.stats
|
|
|
|
from . import colors
|
|
from ._labels import labels
|
|
|
|
|
|
def truncate_text(text, max_len):
|
|
if len(text) > max_len:
|
|
return text[: int(max_len / 2) - 2] + "..." + text[-int(max_len / 2) + 1 :]
|
|
else:
|
|
return text
|
|
|
|
|
|
def monitoring(ind, shap_values, features, feature_names=None, show=True):
|
|
"""Create a SHAP monitoring plot.
|
|
|
|
(Note this function is preliminary and subject to change!!)
|
|
A SHAP monitoring plot is meant to display the behavior of a model
|
|
over time. Often the shap_values given to this plot explain the loss
|
|
of a model, so changes in a feature's impact on the model's loss over
|
|
time can help in monitoring the model's performance.
|
|
|
|
Parameters
|
|
----------
|
|
ind : int
|
|
Index of the feature to plot.
|
|
|
|
shap_values : numpy.array
|
|
Matrix of SHAP values (# samples x # features)
|
|
|
|
features : numpy.array or pandas.DataFrame
|
|
Matrix of feature values (# samples x # features)
|
|
|
|
feature_names : list
|
|
Names of the features (length # features)
|
|
|
|
"""
|
|
if isinstance(features, pd.DataFrame):
|
|
if feature_names is None:
|
|
feature_names = features.columns
|
|
features = features.values
|
|
|
|
num_features = shap_values.shape[1]
|
|
|
|
if feature_names is None:
|
|
feature_names = np.array([labels["FEATURE"] % str(i) for i in range(num_features)])
|
|
|
|
plt.figure(figsize=(10, 3))
|
|
ys = shap_values[:, ind]
|
|
xs = np.arange(len(ys)) # np.linspace(0, 12*2, len(ys))
|
|
|
|
pvals = []
|
|
inc = 50
|
|
for i in range(inc, len(ys) - inc, inc):
|
|
# stat, pval = scipy.stats.mannwhitneyu(v[:i], v[i:], alternative="two-sided")
|
|
_, pval = scipy.stats.ttest_ind(ys[:i], ys[i:])
|
|
pvals.append(pval)
|
|
min_pval = np.min(pvals)
|
|
min_pval_ind = float(np.argmin(pvals) * inc + inc)
|
|
|
|
if min_pval < 0.05 / shap_values.shape[1]:
|
|
plt.axvline(min_pval_ind, linestyle="dashed", color="#666666", alpha=0.2)
|
|
|
|
plt.scatter(xs, ys, s=10, c=features[:, ind], cmap=colors.red_blue)
|
|
|
|
plt.xlabel("Sample index")
|
|
plt.ylabel(truncate_text(feature_names[ind], 30) + "\nSHAP value", size=13)
|
|
plt.gca().xaxis.set_ticks_position("bottom")
|
|
plt.gca().yaxis.set_ticks_position("left")
|
|
plt.gca().spines["right"].set_visible(False)
|
|
plt.gca().spines["top"].set_visible(False)
|
|
cb = plt.colorbar()
|
|
cb.outline.set_visible(False) # type: ignore
|
|
bbox = cb.ax.get_window_extent().transformed(plt.gcf().dpi_scale_trans.inverted())
|
|
cb.ax.set_aspect((bbox.height - 0.7) * 20)
|
|
cb.set_label(truncate_text(feature_names[ind], 30), size=13)
|
|
if show:
|
|
plt.show()
|