def create_merge_group_network_results_workflow(group_list,
data_dir,
subjects_dir,
output_dir,
title='group'):
"""Creates a third-level pipeline to merge the Connectome File Format (CFF) outputs from each group
and combines them into a single CFF file for each group. This version of the third-level pipeline also
concatenates the comma-separated value files for the NetworkX metrics and the connectivity matrices
into single files.
Example
-------
>>> import nipype.workflows.dmri.connectivity.group_connectivity as groupwork
>>> from nipype.testing import example_data
>>> subjects_dir = '.'
>>> data_dir = '.'
>>> output_dir = '.'
>>> group_list = {}
>>> group_list['group1'] = ['subj1', 'subj2']
>>> group_list['group2'] = ['subj3', 'subj4']
>>> l3pipeline = groupwork.create_merge_group_network_results_workflow(group_list, data_dir, subjects_dir, output_dir)
>>> l3pipeline.run() # doctest: +SKIP
Inputs::
group_list: Dictionary of subject lists, keyed by group name
data_dir: Path to the data directory
subjects_dir: Path to the Freesurfer 'subjects' directory
output_dir: Path for the output files
title: String to use as a title for the output merged CFF file (default 'group')
"""
l3infosource = pe.Node(
interface=util.IdentityInterface(fields=['group_id']),
name='l3infosource')
l3infosource.inputs.group_id = group_list.keys()
l3source = pe.Node(nio.DataGrabber(
infields=['group_id'],
outfields=['CFFfiles', 'CSVmetrics', 'CSVmatrices']),
name='l3source')
l3source.inputs.template_args = dict(CFFfiles=[['group_id']],
CSVmetrics=[['group_id']],
CSVmatrices=[['group_id']])
l3source.inputs.template = op.join(output_dir, '%s/%s')
l3source.inputs.field_template = dict(
CFFfiles=op.join(output_dir, '%s/*.cff'),
CSVmetrics=op.join(output_dir, '%s/csv/*.csv'),
CSVmatrices=op.join(output_dir, '%s/cmatrices_csv/*/*.csv'))
l3inputnode = pe.Node(interface=util.IdentityInterface(
fields=['Group_CFFs', 'Group_CSVmetrics', 'Group_CSVmatrices']),
name='l3inputnode')
MergeCNetworks_grp = pe.Node(interface=cmtk.MergeCNetworks(),
name="MergeCNetworks_grp")
MergeCNetworks_grp.inputs.out_file = title
l3datasink = pe.Node(interface=nio.DataSink(), name="l3datasink")
l3datasink.inputs.base_directory = output_dir
l3pipeline = pe.Workflow(name="l3output")
l3pipeline.base_dir = output_dir
l3pipeline.connect([
(l3infosource, l3source, [('group_id', 'group_id')]),
(l3source, l3inputnode, [('CFFfiles', 'Group_CFFs')]),
(l3source, l3inputnode, [('CSVmetrics', 'Group_CSVmetrics')]),
(l3source, l3inputnode, [('CSVmatrices', 'Group_CSVmatrices')]),
])
l3pipeline.connect([(l3inputnode, MergeCNetworks_grp, [('Group_CFFs',
'in_files')])])
l3pipeline.connect([(MergeCNetworks_grp, l3datasink, [('connectome_file',
'@l3output')])])
l3pipeline.connect([(l3inputnode, l3datasink, [
(('Group_CSVmetrics', concatcsv), '@l3output.csvmetrics')
])])
l3pipeline.connect([(l3inputnode, l3datasink, [
(('Group_CSVmatrices', concatcsv), '@l3output.csvmatrices')
])])
return l3pipeline
def create_merge_networks_by_group_workflow(group_list, group_id, data_dir,
subjects_dir, output_dir):
"""Creates a second-level pipeline to merge the Connectome File Format (CFF) outputs from the group-level
MRtrix structural connectivity processing pipeline into a single CFF file for each group.
Example
-------
>>> import nipype.workflows.dmri.connectivity.group_connectivity as groupwork
>>> from nipype.testing import example_data
>>> subjects_dir = '.'
>>> data_dir = '.'
>>> output_dir = '.'
>>> group_list = {}
>>> group_list['group1'] = ['subj1', 'subj2']
>>> group_list['group2'] = ['subj3', 'subj4']
>>> group_id = 'group1'
>>> l2pipeline = groupwork.create_merge_networks_by_group_workflow(group_list, group_id, data_dir, subjects_dir, output_dir)
>>> l2pipeline.run() # doctest: +SKIP
Inputs::
group_list: Dictionary of subject lists, keyed by group name
group_id: String containing the group name
data_dir: Path to the data directory
subjects_dir: Path to the Freesurfer 'subjects' directory
output_dir: Path for the output files
"""
group_infosource = pe.Node(
interface=util.IdentityInterface(fields=['group_id']),
name="group_infosource")
group_infosource.inputs.group_id = group_id
l2infosource = pe.Node(
interface=util.IdentityInterface(fields=['group_id']),
name='l2infosource')
l2source = pe.Node(nio.DataGrabber(infields=['group_id'],
outfields=['CFFfiles']),
name='l2source')
l2source.inputs.template_args = dict(CFFfiles=[['group_id']])
l2source.inputs.template = op.join(output_dir, '%s/cff/*/connectome.cff')
l2source.inputs.base_directory = data_dir
l2inputnode = pe.Node(
interface=util.IdentityInterface(fields=['CFFfiles']),
name='l2inputnode')
MergeCNetworks = pe.Node(interface=cmtk.MergeCNetworks(),
name="MergeCNetworks")
l2datasink = pe.Node(interface=nio.DataSink(), name="l2datasink")
l2datasink.inputs.base_directory = output_dir
l2datasink.inputs.container = group_id
l2pipeline = pe.Workflow(name="l2output")
l2pipeline.base_dir = op.join(output_dir, 'l2output')
l2pipeline.connect([(group_infosource, l2infosource, [('group_id',
'group_id')])])
l2pipeline.connect([
(l2infosource, l2source, [('group_id', 'group_id')]),
(l2source, l2inputnode, [('CFFfiles', 'CFFfiles')]),
])
l2pipeline.connect([(l2inputnode, MergeCNetworks, [('CFFfiles', 'in_files')
])])
l2pipeline.connect([(group_infosource, MergeCNetworks, [('group_id',
'out_file')])])
l2pipeline.connect([(MergeCNetworks, l2datasink, [('connectome_file',
'@l2output')])])
l2pipeline.connect([(group_infosource, l2datasink, [('group_id',
'@group_id')])])
return l2pipeline
def create_merge_network_results_by_group_workflow(group_list, group_id,
data_dir, subjects_dir,
output_dir):
"""Creates a second-level pipeline to merge the Connectome File Format (CFF) outputs from the group-level
MRtrix structural connectivity processing pipeline into a single CFF file for each group.
Example
-------
>>> import nipype.workflows.dmri.connectivity.group_connectivity as groupwork
>>> from nipype.testing import example_data
>>> subjects_dir = '.'
>>> data_dir = '.'
>>> output_dir = '.'
>>> group_list = {}
>>> group_list['group1'] = ['subj1', 'subj2']
>>> group_list['group2'] = ['subj3', 'subj4']
>>> group_id = 'group1'
>>> l2pipeline = groupwork.create_merge_network_results_by_group_workflow(group_list, group_id, data_dir, subjects_dir, output_dir)
>>> l2pipeline.run() # doctest: +SKIP
Inputs::
group_list: Dictionary of subject lists, keyed by group name
group_id: String containing the group name
data_dir: Path to the data directory
subjects_dir: Path to the Freesurfer 'subjects' directory
output_dir: Path for the output files
"""
group_infosource = pe.Node(
interface=util.IdentityInterface(fields=['group_id']),
name="group_infosource")
group_infosource.inputs.group_id = group_id
l2infosource = pe.Node(interface=util.IdentityInterface(fields=[
'group_id',
'merged',
]),
name='l2infosource')
l2source = pe.Node(nio.DataGrabber(infields=['group_id'],
outfields=[
'CFFfiles', 'CSVmatrices',
'CSVfibers', 'CSVnodal', 'CSVglobal'
]),
name='l2source')
l2source.inputs.template_args = dict(CFFfiles=[['group_id']],
CSVmatrices=[['group_id']],
CSVnodal=[['group_id']],
CSVglobal=[['group_id']],
CSVfibers=[['group_id']])
l2source.inputs.base_directory = data_dir
l2source.inputs.template = '%s/%s'
l2source.inputs.field_template = dict(
CFFfiles=op.join(output_dir, '%s/cff/*/connectome.cff'),
CSVmatrices=op.join(output_dir, '%s/cmatrices_csv/*/*.csv'),
CSVnodal=op.join(output_dir, '%s/nxcsv/*/*nodal*.csv'),
CSVglobal=op.join(output_dir, '%s/nxcsv/*/*global*.csv'),
CSVfibers=op.join(output_dir, '%s/fiber_csv/*/*fibers*.csv'))
l2inputnode = pe.Node(interface=util.IdentityInterface(fields=[
'CFFfiles', 'CSVfibers', 'CSVmatrices', 'CSVnodal', 'CSVglobal',
'network_file'
]),
name='l2inputnode')
MergeCNetworks = pe.Node(interface=cmtk.MergeCNetworks(),
name="MergeCNetworks")
l2datasink = pe.Node(interface=nio.DataSink(), name="l2datasink")
l2datasink.inputs.base_directory = output_dir
l2datasink.inputs.container = group_id
l2pipeline = pe.Workflow(name="l2output_" + group_id)
l2pipeline.base_dir = op.join(output_dir, 'l2output')
l2pipeline.connect([(group_infosource, l2infosource, [('group_id',
'group_id')])])
l2pipeline.connect([
(l2infosource, l2source, [('group_id', 'group_id')]),
(l2source, l2inputnode, [('CFFfiles', 'CFFfiles')]),
(l2source, l2inputnode, [('CSVmatrices', 'CSVmatrices')]),
(l2source, l2inputnode, [('CSVnodal', 'CSVnodal')]),
(l2source, l2inputnode, [('CSVglobal', 'CSVglobal')]),
(l2source, l2inputnode, [('CSVfibers', 'CSVfibers')]),
])
l2pipeline.connect([(l2inputnode, MergeCNetworks, [('CFFfiles', 'in_files')
])])
l2pipeline.connect([(group_infosource, MergeCNetworks, [('group_id',
'out_file')])])
l2pipeline.connect([(MergeCNetworks, l2datasink, [('connectome_file',
'@l2output')])])
AddCSVColumn_node = pe.Node(interface=misc.AddCSVColumn(),
name="AddCSVColumn_node")
AddCSVColumn_node.inputs.extra_column_heading = 'group'
AddCSVColumn_global = AddCSVColumn_node.clone(name="AddCSVColumn_global")
AddCSVColumn_matrices = AddCSVColumn_node.clone(
name="AddCSVColumn_matrices")
AddCSVColumn_fibers = AddCSVColumn_node.clone(name="AddCSVColumn_fibers")
concat_csv_interface = Function(input_names=["in_files"],
output_names=["out_name"],
function=concatcsv)
concat_node_csvs = pe.Node(interface=concat_csv_interface,
name='concat_node_csvs')
concat_global_csvs = pe.Node(interface=concat_csv_interface,
name='concat_global_csvs')
concat_matrix_csvs = pe.Node(interface=concat_csv_interface,
name='concat_matrix_csvs')
concat_fiber_csvs = pe.Node(interface=concat_csv_interface,
name='concat_fiber_csvs')
l2pipeline.connect([(l2inputnode, concat_node_csvs, [('CSVnodal',
'in_files')])])
l2pipeline.connect([(concat_node_csvs, AddCSVColumn_node, [('out_name',
'in_file')])])
l2pipeline.connect([(group_infosource, AddCSVColumn_node,
[('group_id', 'extra_field')])])
l2pipeline.connect([(AddCSVColumn_node, l2datasink,
[('csv_file', '@l2output.node_csv')])])
l2pipeline.connect([(group_infosource, l2datasink, [('group_id',
'@group_id')])])
l2pipeline.connect([(l2inputnode, concat_global_csvs, [('CSVglobal',
'in_files')])])
l2pipeline.connect([(concat_global_csvs, AddCSVColumn_global,
[('out_name', 'in_file')])])
l2pipeline.connect([(group_infosource, AddCSVColumn_global,
[('group_id', 'extra_field')])])
l2pipeline.connect([(AddCSVColumn_global, l2datasink,
[('csv_file', '@l2output.global_csv')])])
l2pipeline.connect([(l2inputnode, concat_matrix_csvs, [('CSVmatrices',
'in_files')])])
l2pipeline.connect([(concat_matrix_csvs, AddCSVColumn_matrices,
[('out_name', 'in_file')])])
l2pipeline.connect([(group_infosource, AddCSVColumn_matrices,
[('group_id', 'extra_field')])])
l2pipeline.connect([(AddCSVColumn_matrices, l2datasink,
[('csv_file', '@l2output.cmatrices_csv')])])
l2pipeline.connect([(l2inputnode, concat_fiber_csvs, [('CSVmatrices',
'in_files')])])
l2pipeline.connect([(concat_fiber_csvs, AddCSVColumn_fibers,
[('out_name', 'in_file')])])
l2pipeline.connect([(group_infosource, AddCSVColumn_fibers,
[('group_id', 'extra_field')])])
l2pipeline.connect([(AddCSVColumn_fibers, l2datasink,
[('csv_file', '@l2output.fibers_csv')])])
return l2pipeline
