plot_pairwise_scatter¶

plot_pairwise_scatter(results_a, results_b, method_a, method_b, metric='accuracy', lower_better=False, statistic_tests=True, title=None, figsize=(8, 8), color_palette='tab10')[source]¶

Plot a scatter that compares datasets’ results achieved by two methods.

Parameters:

results_anp.array: Scores (either accuracies or errors) per dataset for the first approach.
results_bnp.array: Scores (either accuracies or errors) per dataset for the second approach.
method_astr: Method name of the first approach.
method_bstr: Method name of the second approach.
metricstr, default = “accuracy”: Metric to be used for the comparison.
lower_betterbool, default = False: If True, lower values are considered better, i.e. errors.
statistic_testsbool, default = True: If True, paired ttest and wilcoxon p-values are shown in the bottom of the plot.
titlestr, default = None: Title to be shown in the top of the plot.
figsizetuple, default = (10, 6): Size of the figure.
color_palettestr, default = “tab10”: Color palette to be used for the plot.

Returns:

figmatplotlib.figure.Figure
axmatplotlib.axes.Axes

Examples

>>> from aeon.visualisation import plot_pairwise_scatter
>>> from aeon.benchmarking.results_loaders import get_estimator_results_as_array
>>> methods = ["InceptionTimeClassifier", "WEASEL-Dilation"]
>>> results = get_estimator_results_as_array(estimators=methods)  
>>> plot = plot_pairwise_scatter(  
...     results[0], methods[0], methods[1])  
>>> plot.show()  
>>> plot.savefig("scatterplot.pdf")  