Just installed Brainstorm 2.2.36, running on Matlab 7.4 and MacOSX 10.4.9. I’m getting the following error messages at startup:
??? Error using ==> find_subdir at 94
ls -R /Applications/Brainstorm/BST_hot Folder/default_database/ failed
Error in ==> browse_study_folder at 101
[subdirs] = find_subdir(dir_name); % alternative to genpath
Error in ==> make_default_database at 279
FILELIST = browse_study_folder(UserDB(iDefault).STUDIES); %update
Error in ==> startup at 379
UserDefault = make_default_database(0); % only check the current database
At the initial startup, Data Manager doesn’t appear. After quitting brainstorm, and attempting startup again, Data Manager does appear, but can’t seem to add new protocols (NEW button in Database Selection and Editor results in the find_subdir at 94 error of above).
Suggestions as to how to fix?
Hi there,
we have limited experience on MacOS, but things should behave just like under Linux I suppose.
The BST startup is attempting to run the following command line:
ls -R /Applications/Brainstorm/BST_hot Folder/default_database/
Do you know why this would fail ?
Please check this command integrity at some terminal prompt on your system (check the path, rights, etc.)
Cheers
Yes, it was path related…
Renamed the main folder to BST_hot, added this folder and its subfolders to Matlab’s search path. All okay.
Thank you!
M-A
That’s right.
BST does not like that pathnames contain spaces (did not realize that you had one when I posted my reply)
Happy brainstorming!
Hi Sylvain,
Thank you for the above replies! I was happily Brainstorming… er at least tutorially Brainstorming when…
Working through with the Database Selector & Editor, I’ve edited the database name (BSTDemo), chose a subject folder (BSTDemo/Subject), and chose a dataset folder (BSTDemo/MEGData). When I hit UPDATE, Matlab shows:
Reference to non-existent field 'StudyName'.
Hitting SAVE, Matlab responds with a swirling blue circle and the message:
Index of element to remove exceeds matrix dimensions.
Nevertheless, went on to import MEG data via the Import Data From window. That process seemed alright with the Brainstorm message window giving reasonable responses. However, after closing the Importing Data From window, the Brainstorm message window says
Refreshing database contents
and Matlab still has a swirling blue circle and shows:
Reference to non-existent field 'StudyName'.
This doesn’t seem like the right way to go, especially since I can’t see that somDemo_av.ds is composed of two trials by clicking on the data (as indicated in the tutorial). Running Matlab 7.4 on MacOSX 10.4. Some advice?
Best,
Mari-Anne
Hmmm… interesting
Maybe the swirling circle in the database manager comes from the fact that the name of the protocole has been altered (therefore this would be a bug from BrainStorm).
Try changing only the Subjects and Studies folders only and hit Update. Before you do that, type brainstorm_reset at the Matlab command prompt and this should flush the corrupted database.
As for the 2 trials, they should appear as 2 separate data files names something like somDemo_data_trial1.mat and somDemo_data_trial2.mat, once imported into BrainStorm.
Hope this will keep you brainstorm happily.
Tried the following (reseting Brainstorm and deleting default databases between each try):
Opened the Default Data protocol in the Database Selection and Editor, changing nothing, UPDATE and SAVE. Everything seems fine. Brainstorm replies with a “Refreshing database… Done.” (Although it does indicate that the data set is empty.)
Opened the Phantom protocol in the Database Selection and Editor, changing nothing. UPDATE gives no response. SAVE leads to “Refreshing database…” from Brainstorm and Matlab has a “Reference to non-existent field ‘StudyName.’”
Opened the Default Data protocol in the Database Selection and Editor. Changed subject (chose folder BSTDemoData/Subject). UPDATE leads to “Refreshing database… Done.” SAVE leads to Matlab’s “Reference to non-existent field ‘StudyName.’”
Opened the Default Data protocol in the Database Selection and Editor. Changed dataset (chose folder BSTDemoData/MEGdata). UPDATE leads to Brainstorm’s “Refreshing database…” SAVE leads to Brainstorm’s “Refreshing database…” and Matlab’s “Reference to non-existent field ‘StudyName.’”
In addition the file selection window called up by changing the dataset has the heading “Select subject (anatomy) folder.” As expected?
Which .m file is associated with the Database Selection and Editor GUI? Is it all in database_manager.m? (Where should ‘StudyName’ be first defined? It looks like there are only 6 .m files that use this… yes?)
Thanks for this detailed report.
So it seems there might be an issue with changing the name and thinsg of the Phantom study right ? Yes database_manager.m is the main script that handles protocoles which are subsequently passed to the data_manager.m
Is this all blocking your progression and can you circumvent this issue and keep on going with the tutorial?
Cheers,
Would love to say that I’ve been able to circumvent and keep going… but unfortunately, most sections require importing some sort of data at which point Brainstorm hangs. (Looks like I can progress through the importing mri tutorial just fine, though.)
It’s not just changing the name, but any type of change to a protocol that has to be saved that leads to some sort of hangup. It seems the error that pops up (“Reference to non-existent field ‘StudyName’”) is in data_manager.m at line 815. 815 reads:
if isnumeric(FileList(Selection).StudyName)
FileList(Selection).StudyName = num2str(FileList(Selection).StudyName);
end
I’ve also got a few “Index of element to remove exceeds matrix dimensions” from Matlab responses when I try to update/save in the database selection editor.
Hmm
What about keeping everything ‘as is originally’ in the protocole without altering anything subsequently ? Things should run quiet smoothly from there.
Please check the path names to your data contain no space or accents.
Cheers
Kept protocols as in original/startup. Path names to data contain no spaces or accents (underscore _ is only exception).
Working with the Phantom protocol. Importing RAW data by associating it to a subject (Montreal Phantom), choosing the Mixed_brainstormstudy.mat and corresponding files for channel locations, etc…
At the end of importing, not so good. It asks whether to transpose a 0 x 0 data matrix. (Seems like it should be a 375x182 matrix.)
Could this be a MacOSX problem ? We haev limited experience on this operating system. Do you have any chance to test run BST under Linux or Windows in your area ? Just to compare and see whether this crashes also outside of Mac ?
thanks for your cooperation.
Hmmm… there must be a Windows machine running Matlab around here somewhere… Will update once I have something.
Thanks,
M-A
Okay, it does appear to be OS related. It runs alright on WinXP!
(Any off-the-top-of-your-head suggestions as to where to start looking for the main sticking points, in case I have a desperate desire to run it on MacOSX?)
Best,
M-A