search

Remi-Gau / NARPS_CPPL

code for the CPPL participation to NARPS
0 stars 0 forks source link

minor bug in iqr_method.m #19

Open marcobarilari opened 5 years ago

marcobarilari commented 5 years ago

https://github.com/Remi-Gau/NARPS_CPPL/blob/4816eb3eb1a5295e2352bb79e799b3308887117a/subfun/iqr_method.m#L27

hello! no idea how to start a new in-line comment for this script so I've found out that opening an 'issue' would be the way.

BTW, in iqr_method there is a bug concerning how the sample size n is calculated and used across the script. It should not change too much.

M.

Remi-Gau commented 5 years ago

For the record. Yes it is good point. This could be an issue in some cases as this way of flagging outliers just pulls over all the data from all runs from all subjects so it is mixing within and between subjects variability. Makes me realize that this function can be better refactored too.