def local_random_affine_transformations(ds,
distort_seeds,
distort_neighbor,
space,
scale_fac=100,
shift_fac=10):
"""Distort a dataset in the local neighborhood of selected features.
This function is similar to ``random_affine_transformation()``, but applies
multiple random affine transformations to a spatially constraint local
neighborhood.
Parameters
----------
ds : Dataset
The to be transformed/distorted dataset.
distort_seeds : list(int)
This a sequence of feature ids (corresponding to the input dataset) that
serve as anchor to determine the local neighborhood for a distortion. The
number of seeds also determines the number of different local distortions
that are going to be applied.
distort_neighbor : callable
And object that when called with a coordinate generates a sequence of
coordinates that comprise its neighborhood (see e.g. ``Sphere()``).
space : str
Name of the feature attribute of the input dataset that contains the
relevant feature coordinates (e.g. 'voxel_indices').
scale_fac : float
See ``random_affine_transformation()``
shift_fac : float
See ``random_affine_transformation()``
Returns
-------
Dataset
A dataset derived from the input dataset with added local distortions.
"""
# which dataset attributes to aggregate
random_stats = ['random_rotation', 'random_scale', 'random_shift']
kwa = {space: distort_neighbor}
qe = IndexQueryEngine(**kwa)
qe.train(ds)
ds_distorted = ds.copy()
for stat in random_stats:
ds_distorted.a[stat + 's'] = {}
# for each seed region
for seed in distort_seeds:
# select the neighborhood for this seed
# take data from the distorted dataset to avoid
# 'loosing' previous distortions
distort_ids = qe[seed]
ds_d = random_affine_transformation(ds_distorted[:, distort_ids],
scale_fac=scale_fac,
shift_fac=shift_fac)
# recover the distortions stats for this seed
for stat in random_stats:
ds_distorted.a[stat + 's'].value[seed] = ds_d.a[stat].value
# put the freshly distorted data back
ds_distorted.samples[:, distort_ids] = ds_d.samples
return ds_distorted
def _get_trained_queryengines(self, datasets, queryengine, radius, ref_ds):
"""Helper to return trained query engine(s), either list of one or one per each dataset
if queryengine is None then IndexQueryEngine based on radius is created
"""
ndatasets = len(datasets)
if queryengine:
if isinstance(queryengine, (list, tuple)):
queryengines = queryengine
if len(queryengines) != ndatasets:
raise ValueError(
"%d query engines were specified although %d datasets "
"provided" % (len(queryengines), ndatasets))
_shpaldebug("Training provided query engines")
for qe, ds in zip(queryengines, datasets):
qe.train(ds)
else:
queryengine.train(datasets[ref_ds])
queryengines = [queryengine]
else:
_shpaldebug('No custom query engines were provided. Setting up the '
'volumetric query engine on voxel_indices.')
queryengine = IndexQueryEngine(voxel_indices=Sphere(radius))
queryengine.train(datasets[ref_ds])
queryengines = [queryengine]
return queryengines
def _get_trained_queryengines(self, datasets, queryengine, radius, ref_ds):
"""Helper to return trained query engine(s), either list of one or one per each dataset
if queryengine is None then IndexQueryEngine based on radius is created
"""
ndatasets = len(datasets)
if queryengine:
if isinstance(queryengine, (list, tuple)):
queryengines = queryengine
if len(queryengines) != ndatasets:
raise ValueError(
"%d query engines were specified although %d datasets "
"provided" % (len(queryengines), ndatasets))
_shpaldebug("Training provided query engines")
for qe, ds in zip(queryengines, datasets):
qe.train(ds)
else:
queryengine.train(datasets[ref_ds])
queryengines = [queryengine]
else:
_shpaldebug(
'No custom query engines were provided. Setting up the '
'volumetric query engine on voxel_indices.')
queryengine = IndexQueryEngine(voxel_indices=Sphere(radius))
queryengine.train(datasets[ref_ds])
queryengines = [queryengine]
return queryengines
def local_random_affine_transformations(
ds, distort_seeds, distort_neighbor, space, scale_fac=100,
shift_fac=10):
"""Distort a dataset in the local neighborhood of selected features.
This function is similar to ``random_affine_transformation()``, but applies
multiple random affine transformations to a spatially constraint local
neighborhood.
Parameters
----------
ds : Dataset
The to be transformed/distorted dataset.
distort_seeds : list(int)
This a sequence of feature ids (corresponding to the input dataset) that
serve as anchor to determine the local neighborhood for a distortion. The
number of seeds also determines the number of different local distortions
that are going to be applied.
distort_neighbor : callable
And object that when called with a coordinate generates a sequence of
coordinates that comprise its neighborhood (see e.g. ``Sphere()``).
space : str
Name of the feature attribute of the input dataset that contains the
relevant feature coordinates (e.g. 'voxel_indices').
scale_fac : float
See ``random_affine_transformation()``
shift_fac : float
See ``random_affine_transformation()``
Returns
-------
Dataset
A dataset derived from the input dataset with added local distortions.
"""
# which dataset attributes to aggregate
random_stats = ['random_rotation', 'random_scale', 'random_shift']
kwa = {space: distort_neighbor}
qe = IndexQueryEngine(**kwa)
qe.train(ds)
ds_distorted = ds.copy()
for stat in random_stats:
ds_distorted.a[stat + 's'] = {}
# for each seed region
for seed in distort_seeds:
# select the neighborhood for this seed
# take data from the distorted dataset to avoid
# 'loosing' previous distortions
distort_ids = qe[seed]
ds_d = random_affine_transformation(
ds_distorted[:, distort_ids],
scale_fac=scale_fac,
shift_fac=shift_fac)
# recover the distortions stats for this seed
for stat in random_stats:
ds_distorted.a[stat + 's'].value[seed] = ds_d.a[stat].value
# put the freshly distorted data back
ds_distorted.samples[:, distort_ids] = ds_d.samples
return ds_distorted
cnx_tx = 489
toutdir = os.path.join(basedir, 'transformation_matrices', 'iterative_cha_olp4cbp_mappers' +'_' + 'subs-' + str(nsubs) + '_'+ 'radius1-10_radius2-' + str(HYPERALIGNMENT_RADIUS) + '.hdf5.gz')
print(toutdir)
# load nifti as a pymvpa dataset and then use that as ref_ds in the queryengine definition
# mask with data in brainmask so only 170k (size of connectomes) voxels are included
ref_ds = fmri_dataset(os.path.join(helperfiles,'newbrainmask.nii'), mask=os.path.join(helperfiles,'newbrainmask.nii'))
print('Size of brain mask:')
print(str(len(ref_ds.fa.voxel_indices)))
# set searchlight sphere radius
sl_radius = HYPERALIGNMENT_RADIUS
#create query engine
qe = IndexQueryEngine(voxel_indices=Sphere(sl_radius))
qe.train(ref_ds)
# load all subject
nfiles = glob.glob(os.path.join(chamats, '*commonspace_subs*'))
print('Loading participant data from: ')
print(chamats)
mysubs = nfiles[0:nsubs]
# import connectomes into pymvpa dataset, zscore, then add chunks and voxel indices, append to list of datsets
dss = []
for sub in range(len(mysubs)):
ds = mv.Dataset(np.load(mysubs[sub]))
ds.fa['voxel_indices'] = range(ds.shape[1])
#ds.sa['chunks'] = np.repeat(i,cnx_tx)
mv.zscore(ds, chunks_attr=None)
dss.append(ds)
