Hello,
We were using topup command with nipype and noticed that some default values indicated in the command documentation is actually different from values used if the users do not specify these parameters. These were related to the subsampling scheme and parameters associated with it;
For example, if we call topup with only required input without specifying --subsamp and other related parameters, there is discrepancy in what the log file indicate as default values and the values actually used;
command:
topup --config=b02b0.cnf --datain=topup_parameters.txt --imain=mydata.nii.gz --out=out_base --iout=out_corrected.nii.gz --fout=out_field.nii.gz --jacout=out_jac --logout=out_topup.log
partial log:
# sub-sampling scheme, default 1
--subsamp=2,2,2,2,2,1,1,1,1
# FWHM (in mm) of gaussian smoothing kernel, default 8
--fwhm=8,6,4,3,3,2,1,0,0
# Max # of non-linear iterations, default 5
--miter=5,5,5,5,5,10,10,20,20
# (approximate) resolution (in mm) of warp basis for the different sub-sampling levels, default 10
--warpres=20,16,14,12,10,6,4,4,4
# Max # of non-linear iterations, default 5
--miter=5,5,5,5,5,10,10,20,20
We found the discrepancy because newer nipype (>=1.0.3) explicitly include default values in the command if the user does not specify them, and it was changing the output from this command compared to the previous version, where they did not have this behaviour.
I just want to make sure that the actual default values, and not those documented, are actually good ones to use..It looks to be the case in our small test case. If they are, would you say these are bugs in the documentation or...?
I would like to know to help nipype behave in the expected way...
Ami Tsuchida
