[Eeglablist] study organization questions

[Arnaud Delorme]

Hi Agatha,

1. Usually, there are 2 files associated with each dataset: FILENAME.set and FILENAME.fdt (.fdt for float data). Both files are important. The .set file contains meta-information (number of channels, sampling frequency etc...) and the .fdt file contains raw data. Note that it is possible to store the raw data in the .set file (using EEGLAB option menu). However, the advantage of having 2 files is that, when dealing with STUDY, EEGLAB can load data channels really fast (which is not possible when using the .set file) without having to load the full raw data file.

2.Yes, no problem. However, copy all the files before creating the new STUDY. If you use both STUDY and they point to exactly the same files (not copies), then they might conflict with each other. One way to keep both STUDY pointing to the same files is to make sure the second STUDY does not use the design indices that are used in the first STUDY. For example, if you have using 5 design in the first STUDY, for the second one, start defining your real design at design 6 (you will have to add 5 designs before you can define number design 6, but you do not have to "use" them -- meaning compute measures for them). We are working on using MD5 file tag so STUDY do not interact with each other (and also different STUDY and design will be able to reuse precomputed files in other design/STUDY).

3. You should be able to copy your STUDY around. We do it all the time.

Thanks,

Arno

On Jun 26, 2012, at 1:14 PM, Agatha Lenartowicz wrote:

> Hi Arno ~
> 
> A couple study level questions.
> 
> 1. I see a file called FILENAME.set in all of my subjects' directories that has size indicating it's a double of the epoched dataset. What is this file and can I delete it (it's eating up space!)?
> 
> 2. I am going to create a new study, re-using files that were previously used in another study. I want to use different studies b/c I can't easily modify the previous study. When I do this - the new study will make all kinds of changes to the design* files produced. I can sort of try to get around this by giving the subjects new *grp* or *condition* names and this will force the design* files to be distinguishable b/w studies. This is however cumbersome and a bit risky if I forget something along the way. So. Is there an easier way?
> 
> 3. How stable are the study structures to changes in paths. At some point I need to back up and archive my study - and I am worried that if I ever return to it, it'll be corrupt.
> 
> 
> Many many thanks!
> Agatha
> 
> 
> 
> 
> _______________________________________________
> 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