PPI concatenate interaction terms

Hello all!

I’m doing a gPPI analysis with a task that has 4 conditions and 4 runs. There were only 2 conditions per run and each condition is only presented twice across all runs. The counterbalancing was pseudorandom so that we didn’t have 2 of the same condition in a given run.

I’ve made it most of the way through the PPI documentation on AFNI. However I’m stumped on concatenating the interaction terms. Since I do not have all conditions present in each run, how should I concatenate them before deconvolution?

Could I do something like we do in the timing file by putting a * for the rows in the 1D file where a condition was not present for that run? Or is there a different convention I should use?