plots.plot_results

plots.plot_results(df, predicted_probas, dict_model, regions=None)

Visualize model prediction results and feature distributions.

This function creates a comprehensive visualization of model prediction results, including feature heatmaps, cumulative probability plots for different folds, and entropy analysis across channels.

Args: df (pd.DataFrame): DataFrame containing channel data and features. predicted_probas (np.ndarray): Array of predicted probabilities with shape (n_folds, n_channels, n_classes) or (n_channels, n_classes). dict_model (dict): Model dictionary containing metadata including features and class information. regions (iblatlas.BrainRegions, optional): BrainRegions object for region visualization. If None, a new instance is created.

Returns: tuple: A tuple containing: - fig (matplotlib.figure.Figure): The figure object containing all plots. - axs (np.ndarray): Array of matplotlib axes objects.

Note: The function creates a multi-panel figure with feature heatmaps, probability plots for each fold, and entropy analysis. It automatically handles both single-fold and multi-fold prediction arrays.