[Eeglablist] EEGLAB2022 study precomputed files directory and parameters

[Delorme, Arnaud]

> Quick clarification on the contents of the single trial files (*ersp/*spec/*erp files). Are the baseline parameters applied to these data, or do they get applied on the fly?

The baseline is applied to the data for ERP. For ERSP and ITC, the complex spectral estimate are saved and the baseline is applied on the fly. For spectrum, there is no baseline (if you choose to remove the mean, it is done on the fly).

> So if I load up the single trial data, can I manually apply whatever baseline or other transform I want (e.g., on the complex values)? Or are the single trials already processed?

You can, expect for ERP. However, you can choose to compute ERP with no baseline.

Cheers,

Arno

> 
> Thank you ~
> Agatha
> 
> 
>> On Apr 28, 2022, at 6:38 AM, Agatha Lenartowicz <alenarto at g.ucla.edu> wrote:
>> 
>> Thanks Arno! I can work with that. Agatha
>> 
>> On Wed, Apr 27, 2022, 7:20 PM Delorme, Arnaud via eeglablist <eeglablist at sccn.ucsd.edu> wrote:
>> Hi Agatha,
>> 
>> > For a variety of reasons, I will have multiple studies for the same dataset. I do NOT wish to create a single study for this dataset, this is not an option for our work. To do so I am setting the ‘filepath’ variable in std_makedesign.m to my desired directory for a given study (it is correctly set in the STUDY.design.filepath. This variable however is being ignored during the precomputation and the precomputed files are being written in the data file directory (default).
>> > 
>> > Question 1. How do I control where the files are being written for different studies? 
>> 
>> The simplest solution to your problem is to copy the STUDY into a separate folder if you want to have two studies pointing to the same files. Note that under Linux/Mac you can delete the EEGLAB set files in the copied folder and replace them with symbolic links to the other STUDY folder if you are concerned about space. This should still work as 2 separate studies.
>> 
>> I know there used to be an option to specify a folder where the files would be precomputed (which we created at your request in EEGLAB 2014), but this is no longer the case. 
>> The reason we removed the option is that you were one of the only ones to use the option, and it was creating too many problems.
>> 
>> > further issue is that even for the the same study, different designs, I am getting warnings that the files will be overwritten for the current STUDY.
>> 
>> You only need to compute the files once. So yes, if you choose to recompute the files (which you do not need to), the files will be overwritten.
>> 
>> > Question 2. How do I verify that the files that get written are respecting different parameters for different designs within a study - where is this recorded?
>> 
>> All the files are MATLAB files, so you can load them from the command line. The parameters are in a field called “parameters.” For example
>> 
>> >> tmp = load('-mat', 'S01.daterp');
>> >> tmp.parameters
>> 
>> ans =
>> 
>>   1×12 cell array
>> 
>>     {'channels'}    {1×71 cell}    {'interp'}    {1×71 struct}    {'savetrials'}    {'on'}    {'recompute'}    {'on'}    {'fileout'}    {'/data/data/STUD…'}    {'trialinfo'}    {1×784 struct}
>> 
>> Best wishes,
>> 
>> Arno
>> _______________________________________________
>> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
>> To unsubscribe, send an empty email to eeglablist-unsubscribe at sccn.ucsd.edu
>> For digest mode, send an email with the subject "set digest mime" to eeglablist-request at sccn.ucsd.edu
>