plot_best_worst_per_point¶
-
mastml.plot_helper.
plot_best_worst_per_point
(y_true, y_pred_list, savepath, metrics_dict, avg_stats, title='best worst per point', label='target_value')[source]¶ Method to create a parity plot (predicted vs. true values) of the set of best and worst CV scores for each
individual data point.
Args:
y_true: (numpy array), array of true y data
y_pred_list: (list), list of numpy arrays containing predicted y data for each CV split
savepath: (str), path to save plots to
metrics_dict: (dict), dict of scikit-learn metric objects to calculate score of predicted vs. true values
avg_stats: (dict), dict of calculated average metrics over all CV splits
title: (str), title of the best_worst_per_point plot
label: (str), label used for axis labeling
Returns:
None