two edf files of same subject

[garvin2506]

Hi,

sorry for bothering you again, it's more a question than an issue this time:
I have two separate edf & beh files for each subject: same ID, but two different task conditions that were performed in separate runs; the TrialIDs are unique across runs (1:40 for the first edf, 41:80 for the second).

I'm wondering what's the best way to do these analyses in your current framework (sticking closely to your nice vignette). Not surprisingly, treating them as two different subjects leads to a mix-up in the merging process because of the shared ID.

I would be grateful for any ideas!

[jashubbard]

If your goal is to produce separate timecourses (fixations or pupil) for the different sessions, then the most straightforward approach would be to create a session variable in your behavioral file before adding it. Then, when you produce your plots, you can include session as one of the grouping variables (just like Task or Conflict in the vignettes). Since you have different trial numbers for the sessions, then it's pretty straightforward:

beh$session <- 1
beh$session[beh$Trial>40] <- 2
z <- add_behdata(z,beh)

This way, all the averages will be computed within-session and within-subject. Will this accomplish what you need?

[garvin2506]

not quite, because the problem starts already when I run set_index and find_messages:

eyetrial starttime endtime duration ID first_sample Results TrialID

1 1 1 9464 9463 3207 12609644 6463 42
2 1 1 9464 9463 3207 12609644 6463 1
3 1 1 9464 9463 3207 12609644 -795824 42
4 1 1 9464 9463 3207 12609644 -795824 1
5 2 9509 19002 9493 3207 12609644 15999 43
6 2 9509 19002 9493 3207 12609644 15999 2
7 2 9509 19002 9493 3207 12609644 -786288 43
8 2 9509 19002 9493 3207 12609644 -786288 2

Here are sample data with 2 edf files and one .rds file (the combined behavioral data, as in your vignette) of one subject: https://www.dropbox.com/sh/woigudoexzxya5u/AAALFWIw77jpdTTVm8xXRWoja?dl=0

here's the relevant code:
z<- set_index(z,varnames=c('TrialID'), patterns=c('Fixation TrialID[1-84]*'), numeric.only=T)
z <- find_messages(z,varnames=c('Results'), patterns=c('^Results'), timestamp=T)
beh$session <- 1
beh$session[beh$TrialID>39] <- 2
z <- add_behdata(z,bet)