def transform_single_imgs(self, imgs, confounds=None, copy=True,
shelve=False):
"""Apply mask, spatial and temporal preprocessing
Parameters
----------
imgs: 3D/4D Niimg-like object
See http://nilearn.github.io/manipulating_visualizing/manipulating_images.html#niimg.
Images to process. It must boil down to a 4D image with scans
number as last dimension.
confounds: CSV file or array-like, optional
This parameter is passed to signal.clean. Please see the related
documentation for details.
shape: (number of scans, number of confounds)
Returns
-------
region_signals: 2D numpy.ndarray
Signal for each voxel inside the mask.
shape: (number of scans, number of voxels)
"""
# Ignore the mask-computing params: they are not useful and will
# just invalid the cache for no good reason
# target_shape and target_affine are conveyed implicitly in mask_img
params = get_params(self.__class__, self,
ignore=['mask_img', 'mask_args', 'mask_strategy'])
data, _ = shelved_filter_and_mask(self, imgs, params,
confounds=confounds,
copy=copy,
shelve=shelve
)
return data
def params_masker(instance, estimator):
"""Get params of NiftiMasker or MultiNiftiMasker instance
Parameters
----------
instance : NiftiMasker or MultiNiftiMasker called object or instance
"""
params = get_params(estimator, instance)
if isinstance(instance, MultiNiftiMasker):
multi_masker = True
def test_filter_and_mask():
data = np.zeros([20, 30, 40, 5])
mask = np.zeros([20, 30, 40, 2])
mask[10, 15, 20, :] = 1
data_img = nibabel.Nifti1Image(data, np.eye(4))
mask_img = nibabel.Nifti1Image(mask, np.eye(4))
masker = NiftiMasker()
params = get_params(NiftiMasker, masker)
assert_raises_regex(DimensionError, "Data must be a 3D", filter_and_mask,
data_img, mask_img, params)
def test_filter_and_mask():
data = np.zeros([20, 30, 40, 5])
mask = np.ones([20, 30, 40])
data_img = nibabel.Nifti1Image(data, np.eye(4))
mask_img = nibabel.Nifti1Image(mask, np.eye(4))
masker = NiftiMasker()
params = get_params(NiftiMasker, masker)
# Test return_affine = False
data = filter_and_mask(data_img, mask_img, params)
assert_equal(data.shape, (5, 24000))
def test_filter_and_mask():
data = np.zeros([20, 30, 40, 5])
mask = np.zeros([20, 30, 40, 2])
mask[10, 15, 20, :] = 1
data_img = nibabel.Nifti1Image(data, np.eye(4))
mask_img = nibabel.Nifti1Image(mask, np.eye(4))
masker = NiftiMasker()
params = get_params(NiftiMasker, masker)
assert_raises_regex(DimensionError, "Data must be a 3D", filter_and_mask,
data_img, mask_img, params)
def test_filter_and_mask():
data = np.zeros([20, 30, 40, 5])
mask = np.ones([20, 30, 40])
data_img = nibabel.Nifti1Image(data, np.eye(4))
mask_img = nibabel.Nifti1Image(mask, np.eye(4))
masker = NiftiMasker()
params = get_params(NiftiMasker, masker)
# Test return_affine = False
data = filter_and_mask(data_img, mask_img, params)
assert_equal(data.shape, (5, 24000))
def test_filter_and_mask_error():
data = np.zeros([20, 30, 40, 5])
mask = np.zeros([20, 30, 40, 2])
mask[10, 15, 20, :] = 1
data_img = nibabel.Nifti1Image(data, np.eye(4))
mask_img = nibabel.Nifti1Image(mask, np.eye(4))
masker = NiftiMasker()
params = get_params(NiftiMasker, masker)
assert_raises_regex(
DimensionError, "Input data has incompatible dimensionality: "
"Expected dimension is 3D and you provided "
"a 4D image.", filter_and_mask, data_img, mask_img, params)
def test_filter_and_mask_error():
data = np.zeros([20, 30, 40, 5])
mask = np.zeros([20, 30, 40, 2])
mask[10, 15, 20, :] = 1
data_img = nibabel.Nifti1Image(data, np.eye(4))
mask_img = nibabel.Nifti1Image(mask, np.eye(4))
masker = NiftiMasker()
params = get_params(NiftiMasker, masker)
assert_raises_regex(DimensionError,
"Input data has incompatible dimensionality: "
"Expected dimension is 3D and you provided "
"a 4D image.",
filter_and_mask,
data_img, mask_img, params)
def transform_single_imgs(self, imgs, confounds=None):
"""Extract signals from a single 4D niimg.
Parameters
----------
imgs: 3D/4D Niimg-like object
See http://nilearn.github.io/manipulating_images/input_output.html
Images to process. It must boil down to a 4D image with scans
number as last dimension.
confounds: CSV file or array-like, optional
This parameter is passed to signal.clean. Please see the related
documentation for details.
shape: (number of scans, number of confounds)
Returns
-------
region_signals: 2D numpy.ndarray
Signal for each sphere.
shape: (number of scans, number of spheres)
"""
self._check_fitted()
params = get_params(NiftiSpheresMasker, self)
signals, _ = self._cache(
filter_and_extract, ignore=['verbose', 'memory', 'memory_level'])(
# Images
imgs,
_ExtractionFunctor(self.seeds_, self.radius, self.mask_img,
self.allow_overlap, self.dtype),
# Pre-processing
params,
confounds=confounds,
dtype=self.dtype,
# Caching
memory=self.memory,
memory_level=self.memory_level,
# kwargs
verbose=self.verbose)
return signals
def our_transform_single_imgs(self, imgs, confounds=None, copy=True):
"""Apply mask, spatial and temporal preprocessing
Parameters
----------
imgs: 3D/4D Niimg-like object
See http://nilearn.github.io/manipulating_images/input_output.html.
Images to process. It must boil down to a 4D image with scans
number as last dimension.
confounds: CSV file or array-like, optional
This parameter is passed to signal.clean. Please see the related
documentation for details.
shape: (number of scans, number of confounds)
Returns
-------
region_signals: 2D numpy.ndarray
Signal for each voxel inside the mask.
shape: (number of scans, number of voxels)
"""
# Ignore the mask-computing params: they are not useful and will
# just invalid the cache for no good reason
# target_shape and target_affine are conveyed implicitly in mask_img
imgs = check_niimg(imgs)
params = get_params(self.__class__, self,
ignore=['mask_img', 'mask_args', 'mask_strategy'])
data = self._cache(filter_and_mask,
ignore=['verbose', 'memory', 'memory_level',
'copy'],
shelve=self._shelving)(
imgs, self.mask_img_, params,
memory_level=self.memory_level,
memory=self.memory,
verbose=self.verbose,
confounds=confounds,
copy=copy
)
return data
def test_get_params():
b = B()
params_a_in_b = class_inspect.get_params(A, b)
assert params_a_in_b == dict(a=1)
params_a_in_b = class_inspect.get_params(A, b, ignore=['a'])
assert params_a_in_b == {}
def check_embedded_nifti_masker(estimator):
"""Base function for using a masker within a BaseEstimator class
This creates a masker from instance parameters :
- If instance contains a mask image in mask parameter,
we use this image as new masker mask_img, forwarding instance parameters to
new masker : smoothing_fwhm, standardize, detrend, low_pass= high_pass,
t_r, target_affine, target_shape, mask_strategy, mask_args,
- If instance contains a masker in mask parameter, we use a copy of
this masker, overriding all instance masker related parameters.
In all case, we forward system parameters of instance to new masker :
memory, memory_level, verbose, n_jobs
Parameters
----------
instance: object, instance of BaseEstimator
The object that gives us the values of the parameters
multi_subject: boolean
Indicates whether to return a MultiNiftiMasker or a NiftiMasker
(the default is True)
Returns
-------
masker: MultiNiftiMasker or NiftiMasker
New masker
"""
estimator_params = get_params(MultiNiftiMasker, estimator)
mask = getattr(estimator, 'mask', None)
if mask is not None and hasattr(mask, 'mask_img'):
# Creating (Multi)NiftiMasker from provided masker
masker_class = mask.__class__
masker_params = get_params(MultiNiftiMasker, mask)
new_masker_params = masker_params
else:
# Creating (Multi)NiftiMasker
# with parameters extracted from estimator
masker_class = MultiNiftiMasker
new_masker_params = estimator_params
new_masker_params['mask_img'] = mask
# Forwarding technical params
new_masker_params['n_jobs'] = estimator.n_jobs
new_masker_params['memory'] = estimator.memory
new_masker_params['memory_level'] = max(0, estimator.memory_level - 1)
new_masker_params['verbose'] = estimator.verbose
# Raising warning if masker override parameters
conflict_string = ""
for param_key in sorted(estimator_params):
if np.any(new_masker_params[param_key] != estimator_params[param_key]):
conflict_string += (
"Parameter {0} :\n"
" Masker parameter {1}"
" - overriding estimator parameter {2}\n").format(
param_key, new_masker_params[param_key],
estimator_params[param_key])
if conflict_string != "":
warn_str = ("Overriding provided-default estimator parameters with"
" provided masker parameters :\n"
"{0:s}").format(conflict_string)
warnings.warn(warn_str)
masker = masker_class(**new_masker_params)
# Forwarding potential attribute of provided masker
if hasattr(mask, 'mask_img_'):
# Allow free fit of returned mask
masker.mask_img = mask.mask_img_
return masker
def transform_imgs(self,
imgs_list,
confounds=None,
copy=True,
n_jobs=1,
shelve=False):
"""Prepare multi subject data in parallel
Parameters
----------
imgs_list: list of Niimg-like objects
See http://nilearn.github.io/manipulating_visualizing/manipulating_images.html#niimg.
List of imgs file to prepare. One item per subject.
confounds: list of confounds, optional
List of confounds (2D arrays or filenames pointing to CSV
files). Must be of same length than imgs_list.
copy: boolean, optional
If True, guarantees that output array has no memory in common with
input array.
n_jobs: integer, optional
The number of cpus to use to do the computation. -1 means
'all cpus'.
Returns
-------
region_signals: list of 2D numpy.ndarray
List of signal for each element per subject.
shape: list of (number of scans, number of elements)
"""
if not hasattr(self, 'mask_img_'):
raise ValueError(
'It seems that %s has not been fitted. '
'You must call fit() before calling transform().' %
self.__class__.__name__)
target_fov = None
if self.target_affine is None:
# Force resampling on first image
target_fov = 'first'
niimg_iter = _iter_check_niimg(imgs_list,
ensure_ndim=None,
atleast_4d=False,
target_fov=target_fov,
memory=self.memory,
memory_level=self.memory_level,
verbose=self.verbose)
if confounds is None:
confounds = itertools.repeat(None, len(imgs_list))
# Ignore the mask-computing params: they are not useful and will
# just invalidate the cache for no good reason
# target_shape and target_affine are conveyed implicitly in mask_img
params = get_params(
self.__class__,
self,
ignore=['mask_img', 'mask_args', 'mask_strategy', 'copy'])
data = Parallel(n_jobs=n_jobs)(
delayed(shelved_filter_and_mask)(self,
imgs,
params,
confounds=cfs,
copy=copy,
shelve=shelve,
squeeze=True)
for imgs, cfs in izip(niimg_iter, confounds))
return data
def test_get_params():
b = B()
params_a_in_b = class_inspect.get_params(A, b)
assert_equal(params_a_in_b, dict(a=1))
params_a_in_b = class_inspect.get_params(A, b, ignore=['a'])
assert_equal(params_a_in_b, {})
def fit(self, niimgs=None, y=None, confounds=None, connectivity=None):
"""Compute the mask and the components
Parameters
----------
niimgs: list of filenames or NiImages
Data on which the PCA must be calculated. If this is a list,
the affine is considered the same for all.
"""
# Hack to support single-subject data:
if isinstance(niimgs, (basestring, nibabel.Nifti1Image)):
niimgs = [niimgs]
# This is a very incomplete hack, as it won't work right for
# single-subject list of 3D filenames
# First, learn the mask
if not isinstance(self.mask, MultiNiftiMasker):
self.masker_ = MultiNiftiMasker(mask=self.mask,
smoothing_fwhm=self.smoothing_fwhm,
target_affine=self.target_affine,
target_shape=self.target_shape,
low_pass=self.low_pass,
high_pass=self.high_pass,
t_r=self.t_r,
memory=self.memory,
memory_level=self.memory_level)
else:
try:
self.masker_ = clone(self.mask)
except TypeError as e:
# Workaround for a joblib bug: in joblib 0.6, a Memory object
# with cachedir = None cannot be cloned.
masker_memory = self.mask.memory
if masker_memory.cachedir is None:
self.mask.memory = None
self.masker_ = clone(self.mask)
self.mask.memory = masker_memory
self.masker_.memory = Memory(cachedir=None)
else:
# The error was raised for another reason
raise e
for param_name in ['target_affine', 'target_shape',
'smoothing_fwhm', 'low_pass', 'high_pass',
't_r', 'memory', 'memory_level']:
if getattr(self.masker_, param_name) is not None:
warnings.warn('Parameter %s of the masker overriden'
% param_name)
setattr(self.masker_, param_name,
getattr(self, param_name))
if self.masker_.mask is None:
self.masker_.fit(niimgs)
else:
self.masker_.fit()
self.mask_img_ = self.masker_.mask_img_
parameters = get_params(MultiNiftiMasker, self)
# Now compute the covariances
self.covariances_ = Parallel(n_jobs=self.n_jobs, verbose=self.verbose)(
delayed(subject_covariance)(
self.estimator,
niimg,
self.masker_.mask_img_,
parameters,
memory=self.memory,
ref_memory_level=self.memory_level,
confounds=confounds,
connectivity=connectivity,
verbose=self.verbose
)
for niimg in niimgs)
return self
def transform_imgs(self, imgs_list, confounds=None, copy=True, n_jobs=1,
shelve=False):
"""Prepare multi subject data in parallel
Parameters
----------
imgs_list: list of Niimg-like objects
See http://nilearn.github.io/manipulating_visualizing/manipulating_images.html#niimg.
List of imgs file to prepare. One item per subject.
confounds: list of confounds, optional
List of confounds (2D arrays or filenames pointing to CSV
files). Must be of same length than imgs_list.
copy: boolean, optional
If True, guarantees that output array has no memory in common with
input array.
n_jobs: integer, optional
The number of cpus to use to do the computation. -1 means
'all cpus'.
Returns
-------
region_signals: list of 2D numpy.ndarray
List of signal for each element per subject.
shape: list of (number of scans, number of elements)
"""
if not hasattr(self, 'mask_img_'):
raise ValueError('It seems that %s has not been fitted. '
'You must call fit() before calling transform().'
% self.__class__.__name__)
target_fov = None
if self.target_affine is None:
# Force resampling on first image
target_fov = 'first'
niimg_iter = _iter_check_niimg(imgs_list, ensure_ndim=None,
atleast_4d=False,
target_fov=target_fov,
memory=self.memory,
memory_level=self.memory_level,
verbose=self.verbose)
if confounds is None:
confounds = itertools.repeat(None, len(imgs_list))
# Ignore the mask-computing params: they are not useful and will
# just invalidate the cache for no good reason
# target_shape and target_affine are conveyed implicitly in mask_img
params = get_params(self.__class__, self,
ignore=['mask_img', 'mask_args', 'mask_strategy',
'copy'])
data = Parallel(n_jobs=n_jobs)(
delayed(shelved_filter_and_mask)(self, imgs, params,
confounds=cfs,
copy=copy,
shelve=shelve,
squeeze=True)
for imgs, cfs in izip(niimg_iter, confounds))
return data
def test_get_params():
b = B()
params_a_in_b = class_inspect.get_params(A, b)
assert_equal(params_a_in_b, dict(a=1))
params_a_in_b = class_inspect.get_params(A, b, ignore=['a'])
assert_equal(params_a_in_b, {})