This repo provides several software modules you can use to setup your own BCI experiments. We use these modules in our own experimental setups at Motorlab and we are releasing them so you can get a head start and not have to build your own modules from scratch which would take a long time.

Prerequisites

1. Install Dragonfly

2. Several of the python modules require numpy, scipy, matplotlib, traits, traitsui, pyface and wxPython

3. Install PyQt from here. You need py2.7 version and make sure you install the one that matches your python build (32-bit or 64-bit)

4. On Linux, scripts in tools folder use screen so make sure you have it installed (sudo apt-get install screen)

Installation

Linux

Clone the repository and set the following environment variables in a permanent place like .bashrc and make sure you export them:

• Set BCI_ROOT to the where you placed your BCI folder
• Set BCI_INCLUDE to ${BCI_ROOT}/message_definitions
• Set BCI_COMMON to ${BCI_ROOT}/modules/Common
• Set BCI_MODULES to ${BCI_ROOT}/modules
• Set BCI_CONFIG to ${BCI_ROOT}/app_configs
• Set BCI_DATA to ${BCI_ROOT}/data
• Append ${BCI_INCLUDE} and ${BCI_COMMON}/python to PYTHONPATH environment variable

For reference, see set_env_vars.sh in tools folder.

Windows

Use the installer which will also set all the necessary environment variables.

After Installation

To play with the example configuration in ${BCI_ROOT}/app_configs, run the following scripts from ${BCI_ROOT}/tools in the order shown below. Note that scripts for windows have .bat extensions while scripts for linux have .sh extensions

1. start_message_manager
2. start_quick_logger
3. start_watcher
4. start_app_starter
5. start_appman_gui

When Appman GUI window pops up, select the configuration for your system, click START and the virtual experiment will start running.