def plot_ics(maps3d, affine, output_dir, titles=None,
format='png', cmap=viz.cm.cold_hot, mean_img=None,
report=True,
parameters=None,
**kwargs):
""" Save the ics to image file, and to a report.
Parameters
----------
titles: None, string or list of strings.
Titles to be used for each ICs
format: {'png', 'pdf', 'svg', 'jpg', ...}
The format used to save a preview.
cmap: matplotlib colormap
The colormap to be used for the independant compnents, for
example pylab.cm.hot
report: boolean, optional
If report is True, an html report is saved.
parameters: None or dictionnary, optional
Extra parameters to put in the report.
kwargs:
Extra keyword arguments are passed to plot_map_2d.
"""
if not os.path.exists(output_dir):
os.makedirs(output_dir)
if mean_img is not None:
img = VolumeImg(mean_img, affine=affine, world_space='mine')
try:
img = img.xyz_ordered()
except CompositionError:
# That's fine, we'll just take it in diagonal
pass
kwargs['anat'] = img.get_data()
kwargs['anat_affine'] = img.affine
img_files = list()
maps3d = np.rollaxis(maps3d, -1)
for index, map3d in enumerate(maps3d):
img_base_file = 'map_%02i.%s' % (index, format)
img_file = pjoin(output_dir, img_base_file)
print 'Outputing image for map %2i out of % 2i: %s' \
% (index+1, len(maps3d), img_file)
# XYZ order the images, this should be done in the viz
# code.
img = VolumeImg(map3d, affine=affine, world_space='mine')
try:
img = img.xyz_ordered()
except CompositionError:
# That's fine, we'll just take it in diagonal
pass
map3d = img.get_data()
this_affine = img.affine
x, y, z = viz.find_cut_coords(map3d, activation_threshold=1e-10)
x, y, z = viz.coord_transform(x, y, z, this_affine)
if np.any(map3d != 0):
# Force the colormap to be symetric:
map_max = max(-map3d.min(), map3d.max())
kwargs['vmin'] = -map_max
kwargs['vmax'] = map_max
map3d = np.ma.masked_equal(map3d, 0, copy=False)
if titles is not None:
if isinstance(titles, basestring):
title = titles % locals()
else:
title = titles[index]
else:
title = None
viz.plot_map(map3d, this_affine, (x, y, z), figure=512,
title=title,
cmap=cmap,
**kwargs)
pl.savefig(img_file)
pl.clf()
img_files.append(img_base_file)
if format in ('png', 'jpg'):
report = markup.page()
report.init(title='CanICA report')
report.p(""" CanICA run, %s.""" % time.asctime())
report.h1("Independent components")
report.img(src=img_files)
if parameters is not None:
report.h1("Parameters")
for name, value in parameters.iteritems():
if isinstance(value, list) or isinstance(value, tuple):
report.p(r'<strong>%s</strong>: %s(' %
(name, value.__class__.__name__))
report.ul()
descriptions = list()
for item in value:
description = pprint.pformat(item)
if len(description) > 1000:
description = ('%s...<br> ...%s'
% (description[:200], description[-200:]))
descriptions.append(description)
report.li(descriptions)
report.ul.close()
report.p(')')
else:
description = pprint.pformat(value)
if len(description) > 1500:
description = ('%s...<br> ...%s'
% (description[:500], description[-500:]))
report.p(r'<strong>%s</strong>: %s' %
(name, description))
report_file = pjoin(output_dir, 'canica.html')
file(report_file, 'w').write(str(report))
print 80*'_'
print 'CanICA: report in %s' % report_file
std[std==0] = 1
series /= std[:, np.newaxis]
u, s, v = linalg.svd(series, full_matrices=False)
s[0] = 0
series = np.dot(u*s, v)
return mask, series
mask, series = mem.cache(load_data)(filenames)
FMRI_MAX = np.abs(series).max()
ref_img = as_volume_img(filenames[0], world_space='mni152')
fmri_affine = ref_img.affine
anat_img = VolumeImg(anat_blurred, anat_affine, world_space='mni152')
anat_img = anat_img.resampled_to_img(ref_img)
blurred_data = anat_img.get_data()
def render_fmri(volume):
data = np.zeros(mask.shape)
data[mask] = volume
data *= blurred_data
fmri_src = viz3d.affine_img_src(data, fmri_affine,
name='fMRI')
vol = mlab.pipeline.volume(fmri_src)
# Change the opacity function
from enthought.tvtk.util.ctf import PiecewiseFunction
def plot_ics(maps3d,
affine,
output_dir,
titles=None,
format='png',
cmap=viz.cm.cold_hot,
mean_img=None,
report=True,
parameters=None,
**kwargs):
""" Save the ics to image file, and to a report.
Parameters
----------
titles: None, string or list of strings.
Titles to be used for each ICs
format: {'png', 'pdf', 'svg', 'jpg', ...}
The format used to save a preview.
cmap: matplotlib colormap
The colormap to be used for the independant compnents, for
example pylab.cm.hot
report: boolean, optional
If report is True, an html report is saved.
parameters: None or dictionnary, optional
Extra parameters to put in the report.
kwargs:
Extra keyword arguments are passed to plot_map_2d.
"""
if not os.path.exists(output_dir):
os.makedirs(output_dir)
if mean_img is not None:
img = VolumeImg(mean_img, affine=affine, world_space='mine')
try:
img = img.xyz_ordered()
except CompositionError:
# That's fine, we'll just take it in diagonal
pass
kwargs['anat'] = img.get_data()
kwargs['anat_affine'] = img.affine
img_files = list()
maps3d = np.rollaxis(maps3d, -1)
for index, map3d in enumerate(maps3d):
img_base_file = 'map_%02i.%s' % (index, format)
img_file = pjoin(output_dir, img_base_file)
print 'Outputing image for map %2i out of % 2i: %s' \
% (index+1, len(maps3d), img_file)
# XYZ order the images, this should be done in the viz
# code.
img = VolumeImg(map3d, affine=affine, world_space='mine')
try:
img = img.xyz_ordered()
except CompositionError:
# That's fine, we'll just take it in diagonal
pass
map3d = img.get_data()
this_affine = img.affine
x, y, z = viz.find_cut_coords(map3d, activation_threshold=1e-10)
x, y, z = viz.coord_transform(x, y, z, this_affine)
if np.any(map3d != 0):
# Force the colormap to be symetric:
map_max = max(-map3d.min(), map3d.max())
kwargs['vmin'] = -map_max
kwargs['vmax'] = map_max
map3d = np.ma.masked_equal(map3d, 0, copy=False)
if titles is not None:
if isinstance(titles, basestring):
title = titles % locals()
else:
title = titles[index]
else:
title = None
viz.plot_map(map3d,
this_affine, (x, y, z),
figure=512,
title=title,
cmap=cmap,
**kwargs)
pl.savefig(img_file)
pl.clf()
img_files.append(img_base_file)
if format in ('png', 'jpg'):
report = markup.page()
report.init(title='CanICA report')
report.p(""" CanICA run, %s.""" % time.asctime())
report.h1("Independent components")
report.img(src=img_files)
if parameters is not None:
report.h1("Parameters")
for name, value in parameters.iteritems():
if isinstance(value, list) or isinstance(value, tuple):
report.p(r'<strong>%s</strong>: %s(' %
(name, value.__class__.__name__))
report.ul()
descriptions = list()
for item in value:
description = pprint.pformat(item)
if len(description) > 1000:
description = (
'%s...<br> ...%s' %
(description[:200], description[-200:]))
descriptions.append(description)
report.li(descriptions)
report.ul.close()
report.p(')')
else:
description = pprint.pformat(value)
if len(description) > 1500:
description = ('%s...<br> ...%s' %
(description[:500], description[-500:]))
report.p(r'<strong>%s</strong>: %s' % (name, description))
report_file = pjoin(output_dir, 'canica.html')
file(report_file, 'w').write(str(report))
print 80 * '_'
print 'CanICA: report in %s' % report_file
def _xyz_order(map, affine):
img = VolumeImg(map, affine=affine, world_space='mine')
img = img.xyz_ordered(resample=True, copy=False)
map = img.get_data()
affine = img.affine
return map, affine
