You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This is a clarification question. In the tutorial code, what exactly does "nibs -c WhiteMatter CSF " do? Is it filtering out csf and white matter columns from my fmriprep regressor file? In the raw regressor file I received from fmriprep, the first 4 columns all start with csf: "csf", "csf_derivative1", "csf_derivative1_power2", "csf_power2", and the next 4 columns all start with whitematter: "white_matter", "white_matter_derivative1", "white_matter_power2", "white_matter_derivative1_power2". Does the command have anything to do with those columns?
I have an additional question. fMRIprep output has an Html file that includes a method section that described the entire data preprocessing steps and I can just copy-paste it into my paper. Is there a similar method description paragraph I could refer to for using the nibetaseries package?
Thanks,
Lily
The text was updated successfully, but these errors were encountered:
Hi,
This is a clarification question. In the tutorial code, what exactly does "nibs -c WhiteMatter CSF " do? Is it filtering out csf and white matter columns from my fmriprep regressor file? In the raw regressor file I received from fmriprep, the first 4 columns all start with csf: "csf", "csf_derivative1", "csf_derivative1_power2", "csf_power2", and the next 4 columns all start with whitematter: "white_matter", "white_matter_derivative1", "white_matter_power2", "white_matter_derivative1_power2". Does the command have anything to do with those columns?
I have an additional question. fMRIprep output has an Html file that includes a method section that described the entire data preprocessing steps and I can just copy-paste it into my paper. Is there a similar method description paragraph I could refer to for using the nibetaseries package?
Thanks,
Lily
The text was updated successfully, but these errors were encountered: