Specify Datasink output path

[Alexandre Savio]

Hi,

I'm trying to run a workflow on many files.
It starts with an IdentityInterface which gives subject_id to a DataGrabber.

This DataGrabber passes 'warp' and 'reference' to a MapNode with fsl.InvWarp and name fsl_invwarp.
So far it works.

Now I'm trying to copy the results of fsl_invwarp that are in workflow.base_dir, 4 subfolders below.

I've instantiated a Node (or should it be a MapNode?) with a DataSink.
For this node I specified inputs.base_directory (why not base_dir? like Workflow...), but this results in many fsl_invwarp%d folders in the base_directory.

I would like the DataSink to create one folder named {subject_id} for each subject and inside it then, the fsl_invwarp.

Can I do that with lambda in the workflow connect? Rename Node? The path rename fix? What could be the best option?

The documentation is not very clear.

Thanks!

My kind regards,
Alex

[Alexandre Savio]

Hi again,

Sorry. I had a mistake in the connect parts of the sink, was not using 'in_file' and the DataSink behaviour was strange.

Now my plan is to Rename them.

Thanks!

Cheers,
Alex