There are two ways to get the BBCI Toolbox:
- get a copy of the latest snapshot, or
- checking it out from the GIT repository.
Variant 2 gives the possibility to keep up-to-date with the most recent version, while snapshots will be provided only after larger updates.
In any case, it is important that you do not modify functions within the toolbox. Your changes will get lost after updating, unless you take great care to transfer them. If you need to change toolbox functions, it is recommended that you make a copy of the functions to another folder that you put in higher priority into the matlab search path and modify functions there.
If you want to run the demos of the toolbox you also need to download the demo datasets (sorry: large files).
Download this archive and unzip it at an appropriate place.
The GIT repository is available at github: https://github.com/bbci/bbci_public
mkdir ~/git
cd ~/git
git clone https://github.com/bbci/bbci_public bbci_public
If you plan to work with our Open Source Framework Pyff for Stimulus Presentation or Feedbacks, you need to get the Pyff repository separately:
git clone git://github.com/bbci/pyff
Open the terminal. If you cannot find the terminal either go to your Desktop or open Finder. Then click on Go -> Utilities. Then type in
mkdir ~/git
cd ~/git
git clone https://github.com/bbci/bbci_public bbci_public
Git for Windows can be downloaded from here:
Installing Git for Windows will add git-related options to the context menu of the Windows Explorer (i.e. you will see these options when you right-click on a folder).
After installing Git for Windows, open the folder in which you would like to store the BBCI Toolbox. In this folder, right-click to open the context menu and choose the option 'Git Bash'. A terminal window (bash) will open. In the terminal window type
git clone https://github.com/bbci/bbci_public
Alternatively, you can use the GUI provided by Git for Windows to clone the repository. The GUI appears when you choose Git GUI from the context menu.
There are files in raw data format (as stored by the EEG recording software,
Brain Vision Recorder in this case), and files that have been converted to
Matlab format already. This typically involves already some kind of
preprocessing (e.g., low-pass filtering, subsampling), and the definition of
classes based on markers. You should dedicate one directory for data of the BBCI
Toolbox, e.g., ~/data
under Linux and d:\data
under Windows and unzip the
following archives there, such that subfolders demoRaw
and demoMat
are
created in your data folder.
Let us assume the Matlab variable MyToolboxDir
holds the path to the BBCI
Toolbox. That could be 'd:\git\bbci_public'
under Windows and
'~/git/bbci_public'
under Linux. Furthermore, you should have one dedicated
folder for data. Let's assume this folder is in the Matlab variable
'MyDataDir'
. That could be 'd:\data'
under Windows. For running the demos,
you should have herein the subfolders demoRaw
and demoMat
from above.
Moreover, it is convenient to have a subfolder tmp
in the data directory. Then
you can startup the BBCI Toolbox like this:
>> cd(MyToolboxDir);
>> startup_bbci_toolbox('DataDir', MyDataDir);
If you want to define a different folder for temporary files, you can do it like
this (to define /tmp/
as temporary folder:
>> cd(MyToolboxDir);
>> startup_bbci_toolbox('DataDir', MyDataDir, 'TmpDir','/tmp/');
There are more options in startup_public_bbci
, that should be described here.
Until then, you have to inspect to code to learn about the other possibilities.
To enable all the demos of the toolbox (which are in the folders demos
and
online/demos
, you need to run the convert-demos first (only one time - the
converted data is saved in a subfolder of your demoMat
, see above):
>> demo_convert_ERPSpeller
>> demo_convert_MotorImagery
>> demo_convert_NIRSData.