Suma group analysis: I,T,B boxes labeled as numeric

First I just want to say thanks for the help given with my previous SUMA problem. That said im back with another 8). I ran my analysis for all my individual subjects, creating standard 141 mesh surfaces for all of them for group analysis as I was instructed to do. The niml dsets all look good. I zero-padded them prior to group analysis as it seemed like a good idea ( I still have the non zero-padded ones also). I loaded all my datasets into uber_subject using the 3dMEMA function and its runs with success creating a 1d file that opens up fine. The problem is all the titles within the I,T,B boxes for selecting coefficients and thresholds just says “numeric” instead of the original titles present in the individual niml dsets so I have no idea what is what. My best guess would be that I have to include the titles in the niml dsets somehow despite indexing the proper bricks in 3dMEMA beforehand but I dont really know, which is why I come to ask the AFNI gurus. Thanks very much!

What is the -prefix name in the 3dMEMA command?
Note that the 1D format will not include column labels,
and will indeed show as numeric. It is much better to
use .niml.dset or even .gii (gifti) formats.

  • rick

Awesome, so Im assuming that using ConvertDset to change my 1d files to niml format should do the trick?

Give it a try and see. The question is whether the index
labels will be passed through. Maybe they are not even

For surface data, I try to avoid the 1D format. Aside from
the label issues you are seeing, the read/write operations
are much slower than for niml.dset or GIFTI.

  • rick

Thanks, I had no idea you could just append .niml.dset onto the end of your prefix name to have MEMA output a niml format file instead of a 1d. Label problem is all fixed up now.

Yes, many output formats can be specified that way.
Glad it’s working for you!

  • rick