Run level confusion

[Hierakonpolis]

I'm not entirely sure this should be considered an issue or whether this would be about the documentation, but running fmriprep through docker and the python wrapper, if I try to specify any run level I get this error:

fmriprep-docker: error: argument analysis_level: invalid choice: 'full' (choose from 'participant')
It would be good if there was either some kind of feedback explaining why this is the only available run level in this case, or if this information was at least easy to find in the documentation. Is this a limitation owed to using the wrapper? Are the other run level options deprecated? Would it be advisable to run fmriprep directly writing the docker command rather than using the wrapper? Since there is an option to specify the run level, I would be tempted to assume that at least in some cases, more than one choice would be available.

[effigies]

This is the BIDS App protocol, which says analysis level is required and can be run, session, participant or group. We only implement per-subject processing.

The fmriprep-docker wrapper is a little more forgiving, since it started life as a debug tool.

[tsalo]

'full' is an option is for the --level keyword argument, while 'participant' is the only valid value for the analysis_level positional argument. @Hierakonpolis I think you might have mixed up those two parameters?

[Hierakonpolis]

Ah I see, specifying fmriprep-docker /path/to/bids /path/to/out/folder participant --level full seems to work, while fmriprep-docker /path/to/bids /path/to/out/folder --level full has the --level option treated as analysis_level, returning
fmriprep-docker: error: argument analysis_level: invalid choice: 'full' (choose from 'participant')