def plot_table_avg()

in scripts/plot_results.py [0:0]

• 41 lines of code
• 7 McCabe index (conditional complexity)

def plot_table_avg(table, dirname, file_name, save=True, block=False, fontsize=12):
    table = table["data"]

    fig, axs = plt.subplots(1, 6, figsize=(13, 2.1))
    axs = axs.flat
    width = 0.5
    for id_d, dataset in zip(range(len(table.keys())), sorted(table.keys())):
        models = table[dataset]
        labels = sorted(models.keys())
        pos = np.arange(len(labels))
        model_means = [models[model]['mean']
                       for model in sorted(models.keys())]
        model_stds = [models[model]['std']
                       for model in sorted(models.keys())]
        legends = []
        for id_m in range(len(pos)):
            l, = axs[id_d].bar(pos[id_m], model_means[id_m], 
                        width=width, color=f'C{id_m}',
                        align='center', ecolor='black', 
                        capsize=7, yerr=model_stds[id_m], linewidth=0.1
                        )
            legends.append(labels[id_m])

        axs[id_d].set_title(dataset)
        axs[id_d].set_xticks(pos)
        axs[id_d].set_ylim(bottom=0)

    axs[0].set_ylabel('Test error')
    plt.tight_layout(pad=0)
    plt.subplots_adjust(wspace=0.3, hspace=0.3)
    plt.legend(legends,
            ncol=6,
            loc="lower center",
            bbox_to_anchor=(-2.8, -0.5))

    if save:
        fig_dirname = "figs/"
        os.makedirs(fig_dirname, exist_ok=True)
        models = '_'.join(sorted(models.keys()))
        plt.savefig(fig_dirname + file_name + '_' + models +'.pdf',
                    format='pdf', bbox_inches='tight')
    if block:
        plt.show(block=False)
        input('Press to close')
        plt.close('all')