mismatch number of trials in OAT

[matsvanes]

In osl_change_spm_eeg_data.m, it can occur that Sc.D and Sc.newdata have a different amount of trials. For example because Sc.newdata was created only on the 'good' trials of Sc.D (e.g. see oat_run_first_level_epoched.m). Copying the 'badtrials' and 'events' fields into D2 (line 112 and further) fails because of this mismatch in trial numbers.

The bug can be reproduced by running tutorial OAT 3 in osl_change_eeg_data, line 117

[matsvanes]

See matsvanes@3e77a46 for a workable fix. It's not complete though, as the same issue can arrise when functions other than oat_run_first_level_epoched.m point to osl_change_spm_eeg_data.m

[matsvanes]

@AJQuinn Have you thought of a way to also fix the event info? I could repeat the assignment of the event data to the structure that goes into osl_change_spm_eeg_data in oat_run_first_level_epoched.m to other functions that call osl_change_spm_eeg_data, but there might be a more clever fix.