Nipype DataGrabber: The 'in_file' trait of a <> instance must be an existing file name, but a value of <> <class 'list'> was specified

Question

I am trying to build a Workflow for normalization/coregistration. Here is my DataGrabber

data_grabber = Node(interface=DataGrabber(infields=['subject_id'], 
                                      outfields=['t1', 'flair', 'les','ref']),
               name ='data_grabber')
data_grabber.inputs.base_directory = base_dir
data_grabber.inputs.template = '*'
data_grabber.inputs.sort_filelist = True
data_grabber.inputs.field_template = dict(t1='T1/%s_BL_T1.nii',
                                   flair='FLAIR/%s_BL_FLAIR.nii',
                                   les='Lesions/%s_BL_FLAIR*.nii',
                                    ref='/usr/share/data/fsl-mni152-templates/%s')
data_grabber.inputs.template_args = dict(t1=[['subject_id']],
                                   flair=[['subject_id']],
                                   les=[['subject_id']],
                                    ref=[['MNI152_T1_1mm.nii.gz']])

data_grabber.inputs.subject_id = [217, 221]

When I try with only one subject everything runs without problem, with more subjects I receive an the following error.

TraitError: The 'in_file' trait of a FLIRTInputSpec instance must be an existing file name, but a value of ['/[...]/FLAIR/217_BL_FLAIR.nii', '/[...]/FLAIR/221_BL_FLAIR.nii'] <class 'list'> was specified.
Error setting node input:
Node: flair_norm
input: in_file
results_file: /[...]/data_grabber/result_data_grabber.pklz
value: ['/[...]/FLAIR/217_BL_FLAIR.nii', '/[...]/FLAIR/221_BL_FLAIR.nii']

Answer 1

DataGrabber only accepts a single in_file. You need to put the DataGrabber into an MapNode:

data_grabber = MapNode(
    interface = DataGrabber(
        infields = ['subject_id'],
        outfields = ['t1', 'flair', 'les','ref']
    ),
    iterfields = ['in_file'],
    name = 'data_grabber'
)