Problem with subplots

[jeffensen]

When drawing multiple subplots on a figure and using draw_annotation() for each of the corresponding axis, the annotations increase in height as if the package wants to stack them although they are plot on different axes:

[Helveg]

Hi! Do you have the script you used to create this graph?

[jeffensen]

# plot output
fig, ax = plt.subplots(1,3,figsize=(9,4), layout='constrained', sharey=True)
sns.barplot(ax=ax[0], data=df, x='trial_type', y='HPCF', alpha=1, errorbar='se')
ax[0].set_title('All Subjects', fontsize=fontsize_labels)
ax[0].tick_params(axis='y', labelsize=fontsize_ticks, which='both', length=3, width=0.8, direction='out', left=True)

for i, group in enumerate(group_labels):
    sns.barplot(ax=ax[i+1], data=df[df.group == group], x='trial_type', y='HPCF', color=group_palette[i], alpha=1, errorbar='se')
    ax[i+1].set_title('Group '+str(i+1), fontsize=fontsize_labels)

for a in ax:
    a.set_xlabel('', fontsize=fontsize_labels)
    a.set_xticks([0, 1])
    xticklabels = ['1', '2']
    a.set_xticklabels(xticklabels, fontsize=fontsize_ticks)
    a.set_ylabel('Percent', fontsize=fontsize_labels)

    # show small tick lines on x-axes
    a.tick_params(axis='x', labelsize=fontsize_ticks-2, which='both', length=3, width=0.8, direction='out', bottom=True)

    # plot p-value annotations from rmANOVA above
    annotations = [tuple(xticklabels + ["0.001"])]  # set p-value to 0.001 to draw 3 asterisks from rmANOVA results
    starbars.draw_annotation(annotations, ns_show=False, fontsize=fontsize_labels+8, text_distance=0.005, bar_gap=0.001, tip_length=0.02, h_gap=0.01, ax=a)

ax[0].set_ylim(60,106);