Multi-panels figures

The function cortex.export.plot_panels plots a number of 3d views of a given volume, in the same matplotlib figure. It does that by saving a temporary image for each view, and then aggregating them in the same figure.

The function needs to be run on a system with a display, since it will launch a webgl viewer. The best way to get the expected results is to keep the webgl viewer visible during the process.

The selection of views and the aggregation is controlled by a list of “panels”. Examples of panels can be imported with:

from cortex.export import params_flatmap_lateral_medial from cortex.export import params_occipital_triple_view

import os
import tempfile

import numpy as np
import matplotlib.pyplot as plt

import cortex

subject = 'S1'

create some artificial data

shape = cortex.db.get_xfm(subject, 'identity').shape
data = np.arange(np.prod(shape)).reshape(shape)
volume = cortex.Volume(data, subject=subject, xfmname='identity')

Show examples of multi-panels figures

params = cortex.export.params_flatmap_lateral_medial
cortex.export.plot_panels(volume, **params)
plt.show()

params = cortex.export.params_occipital_triple_view
cortex.export.plot_panels(volume, **params)
plt.show()

List all predefined angles

base_name = os.path.join(tempfile.mkdtemp(), 'fig')
list_angles = list(cortex.export.save_views.angle_view_params.keys())

filenames = cortex.export.save_3d_views(
    volume, base_name=base_name, list_angles=list_angles,
    list_surfaces=['inflated'] * len(list_angles))

for filename, angle in zip(filenames, list_angles):
    plt.imshow(plt.imread(filename))
    plt.axis('off')
    plt.title(angle)
    plt.show()

List all predefined surfaces

base_name = os.path.join(tempfile.mkdtemp(), 'fig')
list_surfaces = list(cortex.export.save_views.unfold_view_params.keys())

filenames = cortex.export.save_3d_views(
    volume, base_name=base_name,
    list_angles=['lateral_pivot'] * len(list_surfaces),
    list_surfaces=list_surfaces)

for filename, surface in zip(filenames, list_surfaces):
    plt.imshow(plt.imread(filename))
    plt.axis('off')
    plt.title(surface)
    plt.show()