def test_fetch_localizer_calculation_task():
local_url = "file://" + datadir
ids = np.asarray(['S%2d' % i for i in range(94)])
ids = ids.view(dtype=[('subject_id', 'S3')])
file_mock.add_csv('cubicwebexport.csv', ids)
file_mock.add_csv('cubicwebexport2.csv', ids)
# Disabled: cannot be tested without actually fetching covariates CSV file
# All subjects
dataset = datasets.fetch_localizer_calculation_task(data_dir=tmpdir,
url=local_url,
verbose=0)
assert_true(isinstance(dataset.ext_vars, np.recarray))
assert_true(isinstance(dataset.cmaps[0], _basestring))
assert_equal(dataset.ext_vars.size, 94)
assert_equal(len(dataset.cmaps), 94)
# 20 subjects
dataset = datasets.fetch_localizer_calculation_task(n_subjects=20,
data_dir=tmpdir,
url=local_url,
verbose=0)
assert_true(isinstance(dataset.ext_vars, np.recarray))
assert_true(isinstance(dataset.cmaps[0], _basestring))
assert_equal(dataset.ext_vars.size, 20)
assert_equal(len(dataset.cmaps), 20)
def test_fetch_localizer_calculation_task():
local_url = "file://" + datadir
ids = np.asarray(['S%2d' % i for i in range(94)])
ids = ids.view(dtype=[('subject_id', 'S3')])
file_mock.add_csv('cubicwebexport.csv', ids)
file_mock.add_csv('cubicwebexport2.csv', ids)
# Disabled: cannot be tested without actually fetching covariates CSV file
# All subjects
dataset = datasets.fetch_localizer_calculation_task(data_dir=tmpdir,
url=local_url,
verbose=0)
assert_true(isinstance(dataset.ext_vars, np.recarray))
assert_true(isinstance(dataset.cmaps[0], _basestring))
assert_equal(dataset.ext_vars.size, 94)
assert_equal(len(dataset.cmaps), 94)
# 20 subjects
dataset = datasets.fetch_localizer_calculation_task(n_subjects=20,
data_dir=tmpdir,
url=local_url,
verbose=0)
assert_true(isinstance(dataset.ext_vars, np.recarray))
assert_true(isinstance(dataset.cmaps[0], _basestring))
assert_equal(dataset.ext_vars.size, 20)
assert_equal(len(dataset.cmaps), 20)
variates. The user can refer to the
`plot_localizer_mass_univariate_methods.py` example to see how to use these.
"""
# Author: Virgile Fritsch, <*****@*****.**>, May. 2014
import numpy as np
import matplotlib.pyplot as plt
from nilearn import datasets
from nilearn.input_data import NiftiMasker
from nilearn.image import get_data
############################################################################
# Load Localizer contrast
n_samples = 20
localizer_dataset = datasets.fetch_localizer_calculation_task(
n_subjects=n_samples)
tested_var = np.ones((n_samples, 1))
############################################################################
# Mask data
nifti_masker = NiftiMasker(smoothing_fwhm=5,
memory='nilearn_cache',
memory_level=1) # cache options
cmap_filenames = localizer_dataset.cmaps
fmri_masked = nifti_masker.fit_transform(cmap_filenames)
############################################################################
# Anova (parametric F-scores)
from sklearn.feature_selection import f_regression
_, pvals_anova = f_regression(fmri_masked, tested_var,
center=False) # do not remove intercept
from nilearn import datasets
from nilearn.plotting import plot_stat_map, show
from nilearn.input_data import NiftiMasker
from sklearn.feature_selection import f_regression
import numpy as np
import matplotlib.pyplot as plt
"""
This example shows how to use the Localizer dataset in a basic analysis.
A standard Anova is performed (massively univariate F-test) and the resulting
Bonferroni-corrected p-values are plotted. We use a calculation task and 20 subjects out of the 94 available.
"""
# Load Localizer contrast
n_samples = 20 # 20 samples
localizer_dataset = datasets.fetch_localizer_calculation_task(
n_subjects=n_samples
) # Fetch calculation task contrast maps from the localizer.
tested_var = np.ones((n_samples, )) # Ones used in anova test
"""
Neuroimaging data are represented in 4 dimensions: 3 spatial dimensions,
and one dimension to index time or trials. Scikit-learn algorithms, on the other hand,
only accept 2-dimensional samples × features matrices (see section 2.3).
Depending on the setting, voxels and time series can be considered as features or samples.
For example, in spatial independent component analysis (ICA), voxels are samples.
The reduction process from 4D-images to feature vectors comes with the
loss of spatial structure (see Figure 1). It however allows to discard
uninformative voxels, such as the ones outside of the brain. Such voxels
that only carry noise and scanner artifacts would reduce SNR and affect
the quality of the estimation. The selected voxels form a brain mask.
Such a mask is often given along with the datasets or can be computed
The Localizer dataset contains many contrasts and subject-related
variates. The user can refer to the
`plot_localizer_mass_univariate_methods.py` example to see how to use these.
"""
# Author: Virgile Fritsch, <*****@*****.**>, May. 2014
import numpy as np
from scipy import linalg
import matplotlib.pyplot as plt
from nilearn import datasets
from nilearn.input_data import NiftiMasker
### Load Localizer contrast ###################################################
n_samples = 20
localizer_dataset = datasets.fetch_localizer_calculation_task(
n_subjects=n_samples)
tested_var = np.ones((n_samples, 1))
### Mask data #################################################################
nifti_masker = NiftiMasker(
smoothing_fwhm=5,
memory='nilearn_cache', memory_level=1) # cache options
cmap_filenames = localizer_dataset.cmaps
fmri_masked = nifti_masker.fit_transform(cmap_filenames)
### Anova (parametric F-scores) ###############################################
from nilearn._utils.fixes import f_regression
_, pvals_anova = f_regression(fmri_masked, tested_var,
center=False) # do not remove intercept
pvals_anova *= fmri_masked.shape[1]
pvals_anova[np.isnan(pvals_anova)] = 1
variates. The user can refer to the
`plot_localizer_mass_univariate_methods.py` example to see how to use these.
"""
# Author: Virgile Fritsch, <*****@*****.**>, May. 2014
import numpy as np
import matplotlib.pyplot as plt
from nilearn import datasets
from nilearn.maskers import NiftiMasker
from nilearn.image import get_data
############################################################################
# Load Localizer contrast
n_samples = 20
localizer_dataset = datasets.fetch_localizer_calculation_task(
n_subjects=n_samples, legacy_format=False)
tested_var = np.ones((n_samples, 1))
############################################################################
# Mask data
nifti_masker = NiftiMasker(smoothing_fwhm=5,
memory='nilearn_cache',
memory_level=1) # cache options
cmap_filenames = localizer_dataset.cmaps
fmri_masked = nifti_masker.fit_transform(cmap_filenames)
############################################################################
# Anova (parametric F-scores)
from sklearn.feature_selection import f_regression
_, pvals_anova = f_regression(fmri_masked, tested_var,
center=False) # do not remove intercept
# -*- coding: utf-8 -*- """ Testing nilearn toolbox """ import numpy as np import matplotlib.pyplot as plt from nilearn import datasets from nilearn.input_data import NiftiMasker data = datasets.fetch_localizer_calculation_task(data_dir='.')
# -*- coding: utf-8 -*-
"""
Testing nilearn toolbox
"""
import numpy as np
import matplotlib.pyplot as plt
from nilearn import datasets
from nilearn.input_data import NiftiMasker
### Load Localizer contrast ###################################################
n_samples = 20
dataset_files = datasets.fetch_localizer_calculation_task(n_subjects=n_samples,
data_dir='.')
tested_var = np.ones((n_samples, 1))
### Mask data #################################################################
nifti_masker = NiftiMasker(
smoothing_fwhm=5,
memory='nilearn_cache', memory_level=1) # cache options
fmri_masked = nifti_masker.fit_transform(dataset_files.cmaps)
### Anova (parametric F-scores) ###############################################
from nilearn._utils.fixes import f_regression
_, pvals_anova = f_regression(fmri_masked, tested_var,
center=False) # do not remove intercept
pvals_anova *= fmri_masked.shape[1]
pvals_anova[np.isnan(pvals_anova)] = 1
pvals_anova[pvals_anova > 1] = 1
neg_log_pvals_anova = - np.log10(pvals_anova)
neg_log_pvals_anova_unmasked = nifti_masker.inverse_transform(
neg_log_pvals_anova)
