def load_example_fmri_dataset(name="1slice", literal=False):
"""Load minimal fMRI dataset that is shipped with PyMVPA."""
from mvpa2.datasets.sources.openfmri import OpenFMRIDataset
from mvpa2.datasets.mri import fmri_dataset
from mvpa2.misc.io import SampleAttributes
basedir = op.join(pymvpa_dataroot, "haxby2001")
mask = {
"1slice": op.join(pymvpa_dataroot, "mask.nii.gz"),
"25mm": op.join(basedir, "sub001", "masks", "25mm", "brain.nii.gz"),
}[name]
if literal:
model = 1
subj = 1
openfmri = OpenFMRIDataset(basedir)
ds = openfmri.get_model_bold_dataset(model, subj, flavor=name, mask=mask, noinfolabel="rest")
# re-imagine the global time_coords of a concatenated time series
# this is only for the purpose of keeping the example data in the
# exact same shape as it has always been. in absolute terms this makes no
# sense as there is no continuous time in this dataset
ds.sa["run_time_coords"] = ds.sa.time_coords
ds.sa["time_coords"] = np.arange(len(ds)) * 2.5
else:
if name == "25mm":
raise ValueError("The 25mm dataset is no longer available with " "numerical labels")
attr = SampleAttributes(op.join(pymvpa_dataroot, "attributes.txt"))
ds = fmri_dataset(
samples=op.join(pymvpa_dataroot, "bold.nii.gz"), targets=attr.targets, chunks=attr.chunks, mask=mask
)
return ds
def load_example_fmri_dataset(name='1slice', literal=False):
"""Load minimal fMRI dataset that is shipped with PyMVPA."""
from mvpa2.datasets.sources.openfmri import OpenFMRIDataset
from mvpa2.datasets.mri import fmri_dataset
from mvpa2.misc.io import SampleAttributes
basedir = pathjoin(pymvpa_dataroot, 'haxby2001')
mask = {'1slice': pathjoin(pymvpa_dataroot, 'mask.nii.gz'),
'25mm': pathjoin(basedir, 'sub001', 'masks', '25mm',
'brain.nii.gz')}[name]
if literal:
model = 1
subj = 1
openfmri = OpenFMRIDataset(basedir)
ds = openfmri.get_model_bold_dataset(model, subj, flavor=name,
mask=mask, noinfolabel='rest')
# re-imagine the global time_coords of a concatenated time series
# this is only for the purpose of keeping the example data in the
# exact same shape as it has always been. in absolute terms this makes no
# sense as there is no continuous time in this dataset
ds.sa['run_time_coords'] = ds.sa.time_coords
ds.sa['time_coords'] = np.arange(len(ds)) * 2.5
else:
if name == '25mm':
raise ValueError("The 25mm dataset is no longer available with "
"numerical labels")
attr = SampleAttributes(pathjoin(pymvpa_dataroot, 'attributes.txt'))
ds = fmri_dataset(samples=pathjoin(pymvpa_dataroot, 'bold.nii.gz'),
targets=attr.targets, chunks=attr.chunks,
mask=mask)
return ds
def load_example_fmri_dataset(name='1slice', literal=False):
"""Load minimal fMRI dataset that is shipped with PyMVPA."""
from mvpa2.datasets.eventrelated import events2sample_attr
from mvpa2.datasets.sources.openfmri import OpenFMRIDataset
from mvpa2.datasets.mri import fmri_dataset
from mvpa2.misc.io import SampleAttributes
basedir = os.path.join(pymvpa_dataroot, 'openfmri')
mask = {'1slice': os.path.join(pymvpa_dataroot, 'mask.nii.gz'),
'25mm': os.path.join(basedir, 'sub001', 'masks', '25mm',
'brain.nii.gz')}[name]
if literal:
model = 1
subj = 1
openfmri = OpenFMRIDataset(basedir)
ds = openfmri.get_model_bold_dataset(model, subj, flavor=name,
mask=mask, noinfolabel='rest')
# re-imagine the global time_coords of a concatenated time series
# this is only for the purpose of keeping the example data in the
# exact same shape as it has always been. in absolute terms this makes no
# sense as there is no continuous time in this dataset
ds.sa['run_time_coords'] = ds.sa.time_coords
ds.sa['time_coords'] = np.arange(len(ds)) * 2.5
else:
if name == '25mm':
raise ValueError("The 25mm dataset is no longer available with "
"numerical labels")
attr = SampleAttributes(os.path.join(pymvpa_dataroot, 'attributes.txt'))
ds = fmri_dataset(samples=os.path.join(pymvpa_dataroot, 'bold.nii.gz'),
targets=attr.targets, chunks=attr.chunks,
mask=mask)
return ds
def load_datadb_tutorial_data(path=os.path.join(
pymvpa_datadbroot, 'tutorial_data', 'tutorial_data', 'data'),
roi='brain', add_fa=None):
"""Loads the block-design demo dataset from PyMVPA dataset DB.
Parameters
----------
path : str
Path of the directory containing the dataset files.
roi : str or int or tuple or None
Region Of Interest to be used for masking the dataset. If a string is
given a corresponding mask image from the demo dataset will be used
(mask_<str>.nii.gz). If an int value is given, the corresponding ROI
is determined from the atlas image (mask_hoc.nii.gz). If a tuple is
provided it may contain int values that a processed as explained
before, but the union of a ROIs is taken to produce the final mask.
If None, no masking is performed.
add_fa : dict
Passed on to the dataset creator function (see fmri_dataset() for
more information).
"""
import nibabel as nb
from mvpa2.datasets.sources.openfmri import OpenFMRIDataset
task = model = subj = 1
dhandle = OpenFMRIDataset(path)
maskpath = os.path.join(path, 'sub001', 'masks', 'orig')
if roi is None:
mask = None
elif isinstance(roi, str):
mask = os.path.join(maskpath, roi + '.nii.gz')
elif isinstance(roi, int):
nimg = nb.load(os.path.join(maskpath, 'hoc.nii.gz'))
tmpmask = nimg.get_data() == roi
mask = nb.Nifti1Image(tmpmask.astype(int), nimg.get_affine(),
nimg.get_header())
elif isinstance(roi, tuple) or isinstance(roi, list):
nimg = nb.load(os.path.join(maskpath, 'hoc.nii.gz'))
if externals.versions['nibabel'] >= '1.2':
img_shape = nimg.shape
else:
img_shape = nimg.get_shape()
tmpmask = np.zeros(img_shape, dtype='bool')
for r in roi:
tmpmask = np.logical_or(tmpmask, nimg.get_data() == r)
mask = nb.Nifti1Image(tmpmask.astype(int), nimg.get_affine(),
nimg.get_header())
elif isinstance(roi, nb.Nifti1Image):
mask=roi
else:
raise ValueError("Got something as mask that I cannot handle.")
ds = dhandle.get_model_bold_dataset(model, subj, mask=mask, add_fa=add_fa,
noinfolabel='rest')
# fixup time_coords to make the impression of a continuous time series
# this is only necessary until we have changed the tutorial to
# show/encourage run-wise processing
ds.sa['time_coords'] = np.linspace(0, (len(ds) * 2.5), len(ds) + 1)[:-1]
return ds
def create_betas_per_trial_with_pymvpa_roni(study_path, subj, conf, mask_name, flavor, TR):
dhandle = OpenFMRIDataset(study_path)
model = 1
task = 1
# Do this for other tasks as well. not only the first
mask_fname = _opj(study_path, "sub{:0>3d}".format(subj), "masks", conf.mvpa_tasks[0], "{}.nii.gz".format(mask_name))
print mask_fname
run_datasets = []
for run_id in dhandle.get_task_bold_run_ids(task)[subj]:
if type(run_id) == str:
continue
# all_events = dhandle.get_bold_run_model(model, subj, run_id)
all_events = get_bold_run_model(dhandle, 2, subj, run_id)
run_events = []
i = 0
for event in all_events:
if event["task"] == task:
event["condition"] = "{}-{}".format(event["condition"], event["id"])
run_events.append(event)
i += 1
# load BOLD data for this run (with masking); add 0-based chunk ID
run_ds = dhandle.get_bold_run_dataset(subj, task, run_id, flavor=flavor, chunks=run_id - 1, mask=mask_fname)
# convert event info into a sample attribute and assign as 'targets'
run_ds.sa.time_coords = run_ds.sa.time_indices * TR
run_ds.sa["targets"] = events2sample_attr(run_events, run_ds.sa.time_coords, noinfolabel="rest")
# additional time series preprocessing can go here
poly_detrend(run_ds, polyord=1, chunks_attr="chunks")
zscore(run_ds, chunks_attr="chunks", param_est=("targets", ["rest"]), dtype="float32")
glm_dataset = fit_event_hrf_model(run_ds, run_events, time_attr="time_coords", condition_attr="condition")
glm_dataset.sa["targets"] = [x[: x.find("-")] for x in glm_dataset.sa.condition]
glm_dataset.sa["id"] = [x[x.find("-") + 1 :] for x in glm_dataset.sa.condition]
glm_dataset.sa.condition = glm_dataset.sa["targets"]
glm_dataset.sa["chunks"] = [run_id - 1] * len(glm_dataset.samples)
# If a trial was dropped (the subject pressed on a button) than the counter trial from the
# other condition should also be dropped
for pair in conf.conditions_to_compare:
cond_bool = np.array([c in pair for c in glm_dataset.sa["condition"]])
sub_dataset = glm_dataset[cond_bool]
c = Counter(sub_dataset.sa.id)
for value in c:
if c[value] < 2:
id_bool = np.array([value in cond_id for cond_id in glm_dataset.sa["id"]])
glm_dataset = glm_dataset[np.bitwise_not(np.logical_and(id_bool, cond_bool))]
run_datasets.append(glm_dataset)
return vstack(run_datasets, 0)
def create_betas_per_run_with_pymvpa(study_path, subj, conf, mask_name, flavor):
of = OpenFMRIDataset(study_path)
mask_fname = _opj(study_path, "sub{:0>3d}".format(subj), "masks", conf.mvpa_tasks[0], "{}.nii.gz".format(mask_name))
ds = of.get_model_bold_dataset(
model_id=1,
subj_id=subj,
flavor=flavor,
mask=mask_fname,
# preproc_img=smooth,
preproc_ds=detrend,
modelfx=fit_event_hrf_model,
time_attr="time_coords",
condition_attr="condition",
)
return ds
def test_openfmri_dataset():
of = OpenFMRIDataset(os.path.join(pymvpa_dataroot, 'openfmri'))
sub_ids = of.get_subj_ids()
assert_equal(sub_ids, [1, 'phantom'])
assert_equal(of.get_scan_properties(), {'TR': '2.5'})
tasks = of.get_task_descriptions()
assert_equal(tasks, {1: 'object viewing'})
task = tasks.keys()[0]
run_ids = of.get_bold_run_ids(sub_ids[0], task)
assert_equal(run_ids, range(1, 13))
task_runs = of.get_task_bold_run_ids(task)
assert_equal(task_runs, {1: range(1, 13)})
orig_attrs = SampleAttributes(os.path.join(pymvpa_dataroot,
'attributes_literal.txt'))
for subj, runs in task_runs.iteritems():
for run in runs:
# load single run
ds = of.get_bold_run_dataset(subj, task, run, flavor='1slice',
mask=os.path.join(pymvpa_dataroot,
'mask.nii.gz'),
add_sa='bold_moest.txt')
# basic shape
assert_equal(len(ds), 121)
assert_equal(ds.nfeatures, 530)
# functional mapper
assert_equal(ds.O.shape, (121, 40, 20, 1))
# additional attributes present
moest = of.get_bold_run_motion_estimates(subj, task, run)
for i in range(6):
moest_attr = 'bold_moest.txt_%i' % (i,)
assert_true(moest_attr in ds.sa)
assert_array_equal(moest[:,i], ds.sa[moest_attr].value)
# check conversion of model into sample attribute
events = of.get_bold_run_model(subj, task, run)
targets = events2sample_attr(events,
ds.sa.time_coords,
noinfolabel='rest')
assert_array_equal(
orig_attrs['targets'][(run - 1) * 121: run * len(ds)], targets)
assert_equal(ds.sa['subj'][0], subj)
# more basic access
motion = of.get_task_bold_attributes(1, 'bold_moest.txt', np.loadtxt)
assert_equal(len(motion), 12) # one per run
# one per subject, per volume, 6 estimates
assert_equal([m.shape for m in motion], [(1, 121, 6)] * 12)
def make_ds(sub, datapath, flavor): of = OpenFMRIDataset(datapath) ds = of.get_model_bold_dataset( model_id=1, subj_id=sub, #ds = of.get_bold_run_dataset(1,1,1, flavor=flavor, mask=_opj( datapath, 'sub%.3i' % sub, 'masks', 'task001_run001', 'grey.nii.gz'), #preproc_img=smooth, #preproc_ds = detrend, #modelfx=fit_event_hrf_model, time_attr='time_coords', condition_attr='condition') for i in np.unique(ds.chunks)[5:]: detrend(ds[ds.chunks == i])
def make_ds(sub, datapath, flavor):
of = OpenFMRIDataset(datapath)
ds = of.get_model_bold_dataset(
model_id=1,
subj_id=sub,
flavor=flavor,
mask=_opj(datapath, 'sub%.3i' % sub, 'masks', 'task001_run001',
'grey.nii.gz'),
#preproc_img=smooth,
preproc_ds=detrend,
modelfx=fit_event_hrf_model,
time_attr='time_coords',
condition_attr='condition')
ds14 = ds[np.array([c in ['G1', 'G4'] for c in ds.sa['condition']])]
ds23 = ds[np.array([c in ['G2', 'G3'] for c in ds.sa['condition']])]
result_dir = _opj(datapath, 'mvpa', 'ds', flavor)
if not os.path.isdir(result_dir):
os.makedirs(result_dir)
print "{:0>3d}-ds14 {},{}".format(sub, ds14.shape, ds14.sa.condition)
print "{:0>3d}-ds23 {},{}".format(sub, ds23.shape, ds23.sa.condition)
h5save(_opj(result_dir, 'sub%.3i_14_hrf.hdf5' % sub), ds14)
h5save(_opj(result_dir, 'sub%.3i_23_hrf.hdf5' % sub), ds23)
def make_ds(sub, datapath, flavor):
of = OpenFMRIDataset(datapath)
ds = of.get_model_bold_dataset(
model_id=1,
subj_id=sub,
flavor=flavor,
mask=_opj(datapath, "sub%.3i" % sub, "masks", "task001_run001", "grey.nii.gz"),
# preproc_img=smooth,
preproc_ds=detrend,
modelfx=fit_event_hrf_model,
time_attr="time_coords",
condition_attr="condition",
)
ds14 = ds[np.array([c in ["G1", "G4"] for c in ds.sa["condition"]])]
ds23 = ds[np.array([c in ["G2", "G3"] for c in ds.sa["condition"]])]
result_dir = _opj(datapath, "mvpa", "ds", flavor)
if not os.path.isdir(result_dir):
os.makedirs(result_dir)
print "{:0>3d}-ds14 {},{}".format(sub, ds14.shape, ds14.sa.condition)
print "{:0>3d}-ds23 {},{}".format(sub, ds23.shape, ds23.sa.condition)
h5save(_opj(result_dir, "sub%.3i_14_hrf.hdf5" % sub), ds14)
h5save(_opj(result_dir, "sub%.3i_23_hrf.hdf5" % sub), ds23)
def create_betas_per_trial_with_pymvpa(study_path, subj, conf, mask_name, flavor, TR):
dhandle = OpenFMRIDataset(study_path)
model = 1
task = 1
# Do this for other tasks as well. not only the first
mask_fname = _opj(study_path, "sub{:0>3d}".format(subj), "masks", conf.mvpa_tasks[0], "{}.nii.gz".format(mask_name))
print mask_fname
run_datasets = []
for run_id in dhandle.get_task_bold_run_ids(task)[subj]:
if type(run_id) == str:
continue
all_events = dhandle.get_bold_run_model(model, subj, run_id)
run_events = []
i = 0
for event in all_events:
if event["task"] == task:
event["condition"] = "{}-{}".format(event["condition"], i)
run_events.append(event)
i += 1
# load BOLD data for this run (with masking); add 0-based chunk ID
run_ds = dhandle.get_bold_run_dataset(subj, task, run_id, flavor=flavor, chunks=run_id - 1, mask=mask_fname)
# convert event info into a sample attribute and assign as 'targets'
run_ds.sa.time_coords = run_ds.sa.time_indices * TR
print run_id
run_ds.sa["targets"] = events2sample_attr(run_events, run_ds.sa.time_coords, noinfolabel="rest")
# additional time series preprocessing can go here
poly_detrend(run_ds, polyord=1, chunks_attr="chunks")
zscore(run_ds, chunks_attr="chunks", param_est=("targets", ["rest"]), dtype="float32")
glm_dataset = fit_event_hrf_model(run_ds, run_events, time_attr="time_coords", condition_attr="condition")
glm_dataset.sa["targets"] = [x[: x.find("-")] for x in glm_dataset.sa.condition]
glm_dataset.sa.condition = glm_dataset.sa["targets"]
glm_dataset.sa["chunks"] = [run_id - 1] * len(glm_dataset.samples)
run_datasets.append(glm_dataset)
return vstack(run_datasets, 0)
def load_tutorial_data(path=None, roi='brain', add_fa=None, flavor=None):
"""Loads the block-design demo dataset from PyMVPA dataset DB.
Parameters
----------
path : str, optional
Path to the directory with the extracted content of the tutorial
data package. This is only necessary for accessing the full resolution
data. The ``1slice``, and ``25mm`` flavors are shipped with PyMVPA
itself, and the path argument is ignored for them. This function also
honors the MVPA_LOCATION_TUTORIAL_DATA environment variable, and the
respective configuration setting.
roi : str or int or tuple or None, optional
Region Of Interest to be used for masking the dataset. If a string is
given a corresponding mask image from the demo dataset will be used
(mask_<str>.nii.gz). If an int value is given, the corresponding ROI
is determined from the atlas image (mask_hoc.nii.gz). If a tuple is
provided it may contain int values that a processed as explained
before, but the union of a ROIs is taken to produce the final mask.
If None, no masking is performed.
add_fa : dict, optional
Passed on to the dataset creator function (see fmri_dataset() for
more information).
flavor: str, optional
Resolution flavor of the data to load. By default, the data is loaded in
its original resolution. The PyMVPA source distribution contains a '25mm'
flavor that has been downsampled to a very coarse resolution and can be
used for quick test execution. Likewise a ``1slice`` flavor is available
that contents a full-resultion single-slice subset of the dataset.
"""
if path is None:
if flavor in ('1slice', '25mm'):
# we know that this part is there
path = op.join(pymvpa_dataroot)
else:
# check config for info, pretend it is in the working dir otherwise
path = mvpa2.cfg.get('location',
'tutorial data',
default=op.curdir)
# we need the haxby2001 portion of the tutorial data
path = op.join(path, 'haxby2001')
import nibabel as nb
from mvpa2.datasets.sources.openfmri import OpenFMRIDataset
model = subj = 1
dhandle = OpenFMRIDataset(path)
if flavor is None:
maskpath = op.join(path, 'sub001', 'masks', 'orig')
else:
maskpath = op.join(path, 'sub001', 'masks', flavor)
if roi is None:
mask = None
elif isinstance(roi, str):
mask = op.join(maskpath, roi + '.nii.gz')
elif isinstance(roi, int):
nimg = nb.load(op.join(maskpath, 'hoc.nii.gz'))
tmpmask = nimg.get_data() == roi
mask = nb.Nifti1Image(tmpmask.astype(int), nimg.get_affine(),
nimg.get_header())
elif isinstance(roi, tuple) or isinstance(roi, list):
nimg = nb.load(op.join(maskpath, 'hoc.nii.gz'))
if externals.versions['nibabel'] >= '1.2':
img_shape = nimg.shape
else:
img_shape = nimg.get_shape()
tmpmask = np.zeros(img_shape, dtype='bool')
for r in roi:
tmpmask = np.logical_or(tmpmask, nimg.get_data() == r)
mask = nb.Nifti1Image(tmpmask.astype(int), nimg.get_affine(),
nimg.get_header())
elif isinstance(roi, nb.Nifti1Image):
mask = roi
else:
raise ValueError("Got something as mask that I cannot handle.")
ds = dhandle.get_model_bold_dataset(model,
subj,
flavor=flavor,
mask=mask,
add_fa=add_fa,
noinfolabel='rest')
# fixup time_coords to make the impression of a continuous time series
# this is only necessary until we have changed the tutorial to
# show/encourage run-wise processing
ds.sa['time_coords'] = np.linspace(0, (len(ds) * 2.5), len(ds) + 1)[:-1]
return ds
def run(args):
import numpy as np
import pylab as pl
from mvpa2.datasets.sources.openfmri import OpenFMRIDataset
from mvpa2.misc.plot import timeseries_boxplot, concat_ts_boxplot_stats
from mvpa2.misc.stats import compute_ts_boxplot_stats
from mvpa2.base import verbose
of = OpenFMRIDataset(args.path)
data = of.get_task_bold_attributes(args.task,
args.estimate_fname,
np.loadtxt,
exclude_subjs=args.exclude_subjs)
segment_sizes = [len(d[d.keys()[0]]) for d in data]
if args.relative is not None:
# recode per-subject estimates wrt their particular reference
ref = {
subj: d[args.relative[1]]
for subj, d in data[args.relative[0]].iteritems()
}
for d in data:
for subj in d:
if subj in d:
d[subj] -= ref[subj]
print subj, d[subj].mean()
# collapse all data into a per-run (subj x vol x estimate) array
data = [np.array(d.values()) for d in data]
# figure setup
pl.figure(figsize=(12, 5))
ax = pl.subplot(211)
plt_props = {
'translation': 'estimate L2-norm in mm',
'rotation': 'estimate L2-norm in deg'
}
def bxplot(stats, label):
stats = concat_ts_boxplot_stats(stats)
# XXX need some way to expose whether there were missing subjects and
# report proper IDs -- for now resort to whining
verbose(
0,
"List of outlier time series follows (if any) [note, subject IDs are enumarations and may differ from dataset subject IDs in case of missing subjects]"
)
for i, run in enumerate([
np.where(np.sum(np.logical_not(o.mask), axis=0))
for o in stats[1]
]):
sids = run[0]
if len(sids):
verbose(0,
"%s r%.3i: %s" % (label, i + 1, [s + 1 for s in sids]))
timeseries_boxplot(stats[0]['median'],
mean=stats[0]['mean'],
std=stats[0]['std'],
n=stats[0]['n'],
min=stats[0]['min'],
max=stats[0]['max'],
p25=stats[0]['p25'],
p75=stats[0]['p75'],
outlierd=stats[1],
segment_sizes=segment_sizes)
pl.title(label)
xp, xl = pl.xticks()
pl.xticks(xp, ['' for i in xl])
pl.xlim((0, len(stats[0]['n'])))
pl.ylabel(plt_props[label])
if args.rad2deg and args.estimate_order == 'rottrans':
convfunc = np.rad2deg
else:
convfunc = lambda x: x
# first three columns
run_stats = [
compute_ts_boxplot_stats(convfunc(d[..., :3]),
outlier_abs_minthresh=args.outlier_minthresh,
outlier_thresh=args.outlier_stdthresh,
aggfx=np.linalg.norm,
greedy_outlier=True) for d in data
]
ax = pl.subplot(211)
if args.estimate_order == 'transrot':
bxplot(run_stats, 'translation')
else:
bxplot(run_stats, 'rotation')
# last three columns
if args.rad2deg and args.estimate_order == 'transrot':
convfunc = np.rad2deg
else:
convfunc = lambda x: x
run_stats = [
compute_ts_boxplot_stats(convfunc(d[..., 3:]),
outlier_abs_minthresh=args.outlier_minthresh,
outlier_thresh=args.outlier_stdthresh,
aggfx=np.linalg.norm,
greedy_outlier=True) for d in data
]
ax = pl.subplot(212)
if args.estimate_order == 'rottrans':
bxplot(run_stats, 'translation')
else:
bxplot(run_stats, 'rotation')
pl.xlabel('time in fMRI volumes')
if args.savefig is None:
pl.show()
else:
pl.savefig(args.savefig)
#!/usr/bin/python
import sys
from os.path import join as _opj
from mvpa2.datasets.sources.openfmri import OpenFMRIDataset
from mvpa2.datasets.eventrelated import fit_event_hrf_model
from mvpa2.base.hdf5 import h5save
from nilearn.image import smooth_img
import nibabel as nb
datapath = 'BASEDIR'
of = OpenFMRIDataset(datapath)
sub = int(sys.argv[1]) + 1
def smooth(img):
# we need to preserve the original header because the smoothing function
# f***s the TR up
nimg = smooth_img(img, fwhm=2.0)
return nb.Nifti1Image(nimg.get_data(),
img.get_affine(),
header=img.get_header())
ds = of.get_model_bold_dataset(
model_id=1,
subj_id=sub,
flavor='dico_bold7Tp1_to_subjbold7Tp1',
# full brain
def run(args):
import numpy as np
import pylab as pl
from mvpa2.datasets.sources.openfmri import OpenFMRIDataset
from mvpa2.misc.plot import timeseries_boxplot, concat_ts_boxplot_stats
from mvpa2.misc.stats import compute_ts_boxplot_stats
from mvpa2.base import verbose
of = OpenFMRIDataset(args.path)
data = of.get_task_bold_attributes(
args.task, args.estimate_fname, np.loadtxt,
exclude_subjs=args.exclude_subjs)
segment_sizes = [len(d[d.keys()[0]]) for d in data]
if not args.relative is None:
# recode per-subject estimates wrt their particular reference
ref = {subj: d[args.relative[1]]
for subj, d in data[args.relative[0]].iteritems()}
for d in data:
for subj in d:
if subj in d:
d[subj] -= ref[subj]
print subj, d[subj].mean()
# collapse all data into a per-run (subj x vol x estimate) array
data = [np.array(d.values()) for d in data]
# figure setup
pl.figure(figsize=(12, 5))
ax = pl.subplot(211)
plt_props = {
'translation': 'estimate L2-norm in mm',
'rotation': 'estimate L2-norm in deg'
}
def bxplot(stats, label):
stats = concat_ts_boxplot_stats(stats)
# XXX need some way to expose whether there were missing subjects and
# report proper IDs -- for now resort to whining
verbose(0, "List of outlier time series follows (if any) [note, subject IDs are enumarations and may differ from dataset subject IDs in case of missing subjects]")
for i, run in enumerate([np.where(np.sum(np.logical_not(o.mask), axis=0))
for o in stats[1]]):
sids = run[0]
if len(sids):
verbose(0, "%s r%.3i: %s"
% (label, i + 1, [s + 1 for s in sids]))
timeseries_boxplot(stats[0]['median'],
mean=stats[0]['mean'], std=stats[0]['std'], n=stats[0]['n'],
min=stats[0]['min'], max=stats[0]['max'],
p25=stats[0]['p25'], p75=stats[0]['p75'],
outlierd=stats[1], segment_sizes=segment_sizes)
pl.title(label)
xp, xl = pl.xticks()
pl.xticks(xp, ['' for i in xl])
pl.xlim((0, len(stats[0]['n'])))
pl.ylabel(plt_props[label])
if args.rad2deg and args.estimate_order == 'rottrans':
convfunc = np.rad2deg
else:
convfunc = lambda x: x
# first three columns
run_stats = [compute_ts_boxplot_stats(
convfunc(d[...,:3]),
outlier_abs_minthresh=args.outlier_minthresh,
outlier_thresh=args.outlier_stdthresh,
aggfx=np.linalg.norm,
greedy_outlier=True)
for d in data]
ax = pl.subplot(211)
if args.estimate_order == 'transrot':
bxplot(run_stats, 'translation')
else:
bxplot(run_stats, 'rotation')
# last three columns
if args.rad2deg and args.estimate_order == 'transrot':
convfunc = np.rad2deg
else:
convfunc = lambda x: x
run_stats = [compute_ts_boxplot_stats(
convfunc(d[...,3:]),
outlier_abs_minthresh=args.outlier_minthresh,
outlier_thresh=args.outlier_stdthresh,
aggfx=np.linalg.norm,
greedy_outlier=True)
for d in data]
ax = pl.subplot(212)
if args.estimate_order == 'rottrans':
bxplot(run_stats, 'translation')
else:
bxplot(run_stats, 'rotation')
pl.xlabel('time in fMRI volumes')
if args.savefig is None:
pl.show()
else:
pl.savefig(args.savefig)
def run(args):
from mvpa2.base.hdf5 import h5save
ds = None
vol_attr = dict()
if args.add_vol_attr is not None:
# XXX add a way to use the mapper of an existing dataset to
# add a volume attribute without having to load the entire
# mri data again
vol_attr = dict(args.add_vol_attr)
if not len(args.add_vol_attr) == len(vol_attr):
warning("--vol-attr option with duplicate attribute name: " "check arguments!")
verbose(2, "Prepare to add volumetric feature attributes: %s" % vol_attr)
if args.txt_data is not None:
verbose(1, "Load data from TXT file '%s'" % args.txt_data)
samples = _load_from_txt(args.txt_data)
ds = Dataset(samples)
elif args.npy_data is not None:
verbose(1, "Load data from NPY file '%s'" % args.npy_data)
samples = _load_from_npy(args.npy_data)
ds = Dataset(samples)
elif args.mri_data is not None:
verbose(1, "Load data from MRI image(s) %s" % args.mri_data)
from mvpa2.datasets.mri import fmri_dataset
ds = fmri_dataset(args.mri_data, mask=args.mask, add_fa=vol_attr)
elif args.openfmri_modelbold is not None:
verbose(1, "Load data from OpenFMRI model specification %s" % args.openfmri_modelbold)
if not len(args.openfmri_modelbold[3]):
args.openfmri_modelbold[3] = None
# load openfmri dataset
from mvpa2.datasets.sources.openfmri import OpenFMRIDataset
of = OpenFMRIDataset(args.openfmri_modelbold[0])
ds = of.get_model_bold_dataset(
int(args.openfmri_modelbold[1]),
int(args.openfmri_modelbold[2]),
flavor=args.openfmri_modelbold[3],
mask=args.mask,
add_fa=vol_attr,
add_sa=args.add_fsl_mcpar,
)
if ds is None:
if args.data is None:
raise RuntimeError("no data source specific")
else:
ds = hdf2ds(args.data)[0]
else:
if args.data is not None:
verbose(1, "ignoring dataset input in favor of other data source -- remove either one to disambiguate")
# act on all attribute options
ds = process_common_dsattr_opts(ds, args)
if args.openfmri_modelbold is None and args.add_fsl_mcpar is not None:
from mvpa2.misc.fsl.base import McFlirtParams
mc_par = McFlirtParams(args.add_fsl_mcpar)
for param in mc_par:
verbose(2, "Add motion regressor as sample attribute '%s'" % ("mc_" + param))
ds.sa["mc_" + param] = mc_par[param]
verbose(3, "Dataset summary %s" % (ds.summary()))
# and store
outfilename = args.output
if not outfilename.endswith(".hdf5"):
outfilename += ".hdf5"
verbose(1, "Save dataset to '%s'" % outfilename)
h5save(outfilename, ds, mkdir=True, compression=args.hdf5_compression)
def load_tutorial_data(path=None, roi="brain", add_fa=None, flavor=None):
"""Loads the block-design demo dataset from PyMVPA dataset DB.
Parameters
----------
path : str, optional
Path to the directory with the extracted content of the tutorial
data package. This is only necessary for accessing the full resolution
data. The ``1slice``, and ``25mm`` flavors are shipped with PyMVPA
itself, and the path argument is ignored for them. This function also
honors the MVPA_LOCATION_TUTORIAL_DATA environment variable, and the
respective configuration setting.
roi : str or int or tuple or None, optional
Region Of Interest to be used for masking the dataset. If a string is
given a corresponding mask image from the demo dataset will be used
(mask_<str>.nii.gz). If an int value is given, the corresponding ROI
is determined from the atlas image (mask_hoc.nii.gz). If a tuple is
provided it may contain int values that a processed as explained
before, but the union of a ROIs is taken to produce the final mask.
If None, no masking is performed.
add_fa : dict, optional
Passed on to the dataset creator function (see fmri_dataset() for
more information).
flavor: str, optional
Resolution flavor of the data to load. By default, the data is loaded in
its original resolution. The PyMVPA source distribution contains a '25mm'
flavor that has been downsampled to a very coarse resolution and can be
used for quick test execution. Likewise a ``1slice`` flavor is available
that contents a full-resultion single-slice subset of the dataset.
"""
if path is None:
if flavor in ("1slice", "25mm"):
# we know that this part is there
path = op.join(pymvpa_dataroot)
else:
# check config for info, pretend it is in the working dir otherwise
path = mvpa2.cfg.get("location", "tutorial data", default=op.curdir)
# we need the haxby2001 portion of the tutorial data
path = op.join(path, "haxby2001")
import nibabel as nb
from mvpa2.datasets.sources.openfmri import OpenFMRIDataset
model = subj = 1
dhandle = OpenFMRIDataset(path)
if flavor is None:
maskpath = op.join(path, "sub001", "masks", "orig")
else:
maskpath = op.join(path, "sub001", "masks", flavor)
if roi is None:
mask = None
elif isinstance(roi, str):
mask = op.join(maskpath, roi + ".nii.gz")
elif isinstance(roi, int):
nimg = nb.load(op.join(maskpath, "hoc.nii.gz"))
tmpmask = nimg.get_data() == roi
mask = nb.Nifti1Image(tmpmask.astype(int), nimg.get_affine(), nimg.get_header())
elif isinstance(roi, tuple) or isinstance(roi, list):
nimg = nb.load(op.join(maskpath, "hoc.nii.gz"))
if externals.versions["nibabel"] >= "1.2":
img_shape = nimg.shape
else:
img_shape = nimg.get_shape()
tmpmask = np.zeros(img_shape, dtype="bool")
for r in roi:
tmpmask = np.logical_or(tmpmask, nimg.get_data() == r)
mask = nb.Nifti1Image(tmpmask.astype(int), nimg.get_affine(), nimg.get_header())
elif isinstance(roi, nb.Nifti1Image):
mask = roi
else:
raise ValueError("Got something as mask that I cannot handle.")
ds = dhandle.get_model_bold_dataset(model, subj, flavor=flavor, mask=mask, add_fa=add_fa, noinfolabel="rest")
# fixup time_coords to make the impression of a continuous time series
# this is only necessary until we have changed the tutorial to
# show/encourage run-wise processing
ds.sa["time_coords"] = np.linspace(0, (len(ds) * 2.5), len(ds) + 1)[:-1]
return ds
#!/usr/bin/python
import sys
from os.path import join as _opj
from mvpa2.datasets.sources.openfmri import OpenFMRIDataset
from mvpa2.datasets.eventrelated import fit_event_hrf_model
from mvpa2.base.hdf5 import h5save
from nilearn.image import smooth_img
import nibabel as nb
datapath = 'BASEDIR'
of = OpenFMRIDataset(datapath)
sub = int(sys.argv[1]) + 1
def smooth(img):
# we need to preserve the original header because the smoothing function
# f***s the TR up
nimg = smooth_img(img, fwhm=2.0)
return nb.Nifti1Image(nimg.get_data(),
img.get_affine(),
header=img.get_header())
ds = of.get_model_bold_dataset(
model_id=1, subj_id=sub,
flavor='dico_bold7Tp1_to_subjbold7Tp1',
# full brain
mask=_opj(
datapath, 'sub%.3i' % sub, 'templates', 'bold7Tp1',
def run(args):
from mvpa2.base.hdf5 import h5save
ds = None
vol_attr = dict()
if args.add_vol_attr is not None:
# XXX add a way to use the mapper of an existing dataset to
# add a volume attribute without having to load the entire
# mri data again
vol_attr = dict(args.add_vol_attr)
if not len(args.add_vol_attr) == len(vol_attr):
warning("--vol-attr option with duplicate attribute name: "
"check arguments!")
verbose(2,
"Prepare to add volumetric feature attributes: %s" % vol_attr)
if args.txt_data is not None:
verbose(1, "Load data from TXT file '%s'" % args.txt_data)
samples = _load_from_txt(args.txt_data)
ds = Dataset(samples)
elif args.npy_data is not None:
verbose(1, "Load data from NPY file '%s'" % args.npy_data)
samples = _load_from_npy(args.npy_data)
ds = Dataset(samples)
elif args.mri_data is not None:
verbose(1, "Load data from MRI image(s) %s" % args.mri_data)
from mvpa2.datasets.mri import fmri_dataset
ds = fmri_dataset(args.mri_data, mask=args.mask, add_fa=vol_attr)
elif args.openfmri_modelbold is not None:
verbose(
1, "Load data from OpenFMRI model specification %s" %
args.openfmri_modelbold)
if not len(args.openfmri_modelbold[3]):
args.openfmri_modelbold[3] = None
# load openfmri dataset
from mvpa2.datasets.sources.openfmri import OpenFMRIDataset
of = OpenFMRIDataset(args.openfmri_modelbold[0])
ds = of.get_model_bold_dataset(int(args.openfmri_modelbold[1]),
int(args.openfmri_modelbold[2]),
flavor=args.openfmri_modelbold[3],
mask=args.mask,
add_fa=vol_attr,
add_sa=args.add_fsl_mcpar)
if ds is None:
if args.data is None:
raise RuntimeError('no data source specific')
else:
ds = hdf2ds(args.data)[0]
else:
if args.data is not None:
verbose(
1,
'ignoring dataset input in favor of other data source -- remove either one to disambiguate'
)
# act on all attribute options
ds = process_common_dsattr_opts(ds, args)
if args.openfmri_modelbold is None and args.add_fsl_mcpar is not None:
from mvpa2.misc.fsl.base import McFlirtParams
mc_par = McFlirtParams(args.add_fsl_mcpar)
for param in mc_par:
verbose(
2, "Add motion regressor as sample attribute '%s'" %
('mc_' + param))
ds.sa['mc_' + param] = mc_par[param]
verbose(3, "Dataset summary %s" % (ds.summary()))
# and store
outfilename = args.output
if not outfilename.endswith('.hdf5'):
outfilename += '.hdf5'
verbose(1, "Save dataset to '%s'" % outfilename)
h5save(outfilename, ds, mkdir=True, compression=args.hdf5_compression)
def run(args):
import numpy as np
import pylab as pl
from mvpa2.datasets.sources.openfmri import OpenFMRIDataset
from mvpa2.misc.plot import timeseries_boxplot, concat_ts_boxplot_stats
from mvpa2.misc.stats import compute_ts_boxplot_stats
of = OpenFMRIDataset(args.path)
data = of.get_task_bold_attributes(
args.task, args.estimate_fname, np.loadtxt,
exclude_subjs=args.exclude_subjs)
segment_sizes = [len(d[0]) for d in data]
# figure setup
pl.figure(figsize=(12, 5))
ax = pl.subplot(211)
# translation
run_stats = [compute_ts_boxplot_stats(
d[...,:3],
outlier_abs_minthresh=args.outlier_minthresh,
outlier_thresh=args.outlier_stdthresh,
aggfx=np.linalg.norm,
greedy_outlier=True)
for d in data]
stats = concat_ts_boxplot_stats(run_stats)
timeseries_boxplot(stats[0]['median'],
mean=stats[0]['mean'], std=stats[0]['std'], n=stats[0]['n'],
min=stats[0]['min'], max=stats[0]['max'],
p25=stats[0]['p25'], p75=stats[0]['p75'],
outlierd=stats[1], segment_sizes=segment_sizes)
pl.title('translation')
xp, xl = pl.xticks()
pl.xticks(xp, ['' for i in xl])
pl.xlim((0, len(stats[0]['n'])))
#pl.ylim((0,7))
pl.ylabel('estimate L2-norm in mm')
# rotation
ax = pl.subplot(212)
run_stats = [compute_ts_boxplot_stats(
d[...,3:],
outlier_abs_minthresh=args.outlier_minthresh,
outlier_thresh=args.outlier_stdthresh,
aggfx=np.linalg.norm,
greedy_outlier=True)
for d in data]
stats = concat_ts_boxplot_stats(run_stats)
timeseries_boxplot(stats[0]['median'],
mean=stats[0]['mean'], std=stats[0]['std'], n=stats[0]['n'],
min=stats[0]['min'], max=stats[0]['max'],
p25=stats[0]['p25'], p75=stats[0]['p75'],
outlierd=stats[1], segment_sizes=segment_sizes)
pl.xlim((0, len(stats[0]['n'])))
#pl.ylim((0,5))
pl.title('rotation')
pl.ylabel('estimate L2-norm in deg')
pl.xlabel('time in fMRI volumes')
if args.savefig is None:
pl.show()
else:
pl.savefig(args.safefig)
