def test_prepare_abi_connectivity_maps():
from samri.fetch.local import prepare_abi_connectivity_maps
prepare_abi_connectivity_maps(
'Ventral_tegmental_area',
invert_lr_experiments=[
"127651139",
"127796728",
"127798146",
"127867804",
"156314762",
"160539283",
"160540751",
"165975096",
"166054222",
"171021829",
"175736945",
"278178382",
"292958638",
"301062306",
"304337288",
],
)
def abi_connectivity_map(
identifier,
exclude_experiments=[],
keep_work=False,
mask='/usr/share/mouse-brain-atlases/dsurqec_200micron_mask.nii',
prepare_root='/var/tmp/{user}/samri/abi_connectivity/',
prepare_subdirs='sub-{experiment}/ses-1/anat/sub-{experiment}_ses-1_desc-cope.nii.gz',
save_as_cope='',
save_as_zstat='',
save_as_tstat='',
tmp_dir='/var/tmp/{user}/samri/abi_connectivity/l2',
abi_data_root='/usr/share/ABI-connectivity-data/',
invert_lr_experiments=[],
):
"""Create statistical summary maps (any subset of COPE, t-statistics, and z-statistics) for an ABI connectivity identifier.
Parameters
----------
identifier : str
Experiment set identifier which corresponds to the data set paths from the ABI-connectivity-data package.
exclude_experiments : list of str, optional
List of strings, each string 9 characters long, identifying which experiments should be excluded from the modelling.
keep_work : bool, optional
Whether to keep the work files (including the prepared ABI experiment data, the work directory, results directory, and crash directory of the model).
mask : string, optional
Path to a NIfTI file containing ones and zeroes and specifying the mask for the modelling workflow.
It is important that this data is in the same template space as the ABI-connectivity-data package, which is the DSURQEC space [1]_ .
prepare_root : string, optional
Python-formattable string, under which the prepared (e.g. flipped) data from ABI experiments is to be saved.
Generally this should be a temporal and new path, containing either "{user}" or located under the user's home directory, in order to avoid race conditions between users.
It will be deleted without request for confirmation if this function is executed with `keep_work=False` (which is the default).
prepare_subdirs : string, optional
Python-formattable string, containing "{experiment}" according to which the prepared (e.g. flipped) data from ABI experiments is to be organized inside the `prepare_root` directory.
save_as_cope : str, optional
Path under which to save the COPE result of the modelling.
save_as_tstat : str, optional
Path under which to save the t-statistic result of the modelling.
save_as_zstat : str, optional
Path under which to save the z-statistic result of the modelling.
tmp_dir : string, optional
Temporary directory inside which to execute the modelling workflow.
Generally this should be a temporal and new path, containing either "{user}" or located under the user's home directory, in order to avoid race conditions between users.
abi_data_root : str, optional
Root path for the ABI-connectivity-data package installation on the current machine.
invert_lr_experiments : list of str, optional
List of strings, each string 9 characters long, identifying which experiments need to be inverted with respect to the left-right orientation.
Notes
-----
If neither of the `save_as_cope`, `save_as_tstat`, and `save_as_zstat` parameters are specified, all of the results are saved in the current work directory as `{identifier}_{statistic}.nii.gz`.
References
----------
.. [1] H. I. Ioanas and M. Marks and M. F. Yanik and M. Rudin "An Optimized Registration Workflow and Standard Geometric Space for Small Animal Brain Imaging" https://doi.org/10.1101/619650
"""
# Prepend user name to SAMRI temp directories to prevent users from overwriting each other's work.
current_user = getpass.getuser()
prepare_root = prepare_root.format(user=current_user)
tmp_dir = tmp_dir.format(user=current_user)
reposit_path = path.join(prepare_root, '{identifier}', prepare_subdirs)
prepare_abi_connectivity_maps(
identifier,
abi_data_root=abi_data_root,
reposit_path=reposit_path,
invert_lr_experiments=invert_lr_experiments,
)
prepare_root = prepare_root.format(identifier=identifier)
glm.l2_common_effect(
prepare_root,
workflow_name=identifier,
mask=mask,
n_jobs_percentage=.33,
out_base=tmp_dir,
exclude={'subject': exclude_experiments},
run_mode='fe',
keep_work=keep_work,
)
if save_as_zstat:
zstat_path = path.join(tmp_dir, identifier, '_zstat.nii.gz')
shutil.copyfile(zstat_path, save_as_zstat)
if save_as_tstat:
tstat_path = path.join(tmp_dir, identifier, '_tstat.nii.gz')
shutil.copyfile(tstat_path, save_as_tstat)
if save_as_cope:
cope_path = path.join(tmp_dir, identifier, '_cope.nii.gz')
shutil.copyfile(cope_path, save_as_cope)
if not any([save_as_zstat, save_as_tstat, save_as_cope]):
zstat_path = path.join(tmp_dir, identifier, '_zstat.nii.gz')
shutil.copyfile(zstat_path, '{identifier}_zstat.nii.gz')
tstat_path = path.join(tmp_dir, identifier, '_tstat.nii.gz')
shutil.copyfile(tstat_path, '{identifier}_tstat.nii.gz')
cope_path = path.join(tmp_dir, identifier, '_cope.nii.gz')
shutil.copyfile(cope_path, '{identifier}_cope.nii.gz')
if not keep_work:
results_dir = path.join(tmp_dir, identifier)
shutil.rmtree(results_dir)
shutil.rmtree(prepare_root)