As of 2017-05-12, the following experiments are supported:

• FR1
• catFR1
• FR3
• catFR3
• PS4/FR5
• PS4/catFR5
• FR5/REC1
• catFR5/REC1
• PAL1
• PS4/PAL5
• PAL5

• Ensure permissions are correct: sudo chown -R exp /Users/exp/miniconda2/lib
• Check if Xcode needs to be updated (you'll need Xcode version 8.x) by opening the App Store.
• If homebrew is not installed, install it
• Install portaudio: brew install portaudio

Clone the v3.1 branch of RAMControl:

$ git clone https://github.com/ramdarpaprojectorg/RAMControl.git -b v3.1 ~/RAM_3.1

Then cd ~/RAM_3.1 and run:

./install.sh

Please see the CHANGELOG.md file for instructions.

The catFR practice list differs from that of FR to avoid reuse of some of the words. This change was initially made in System 2 revision 165 and has been used in System 3 since version 3.0.0.

Tasks can be started in debug mode to diagnose problems. This requires opening a terminal, navigating to the directory with the task code, and running with:

$ python run.py -d

(see also --help to see other options). This will start PyEPL in windowed mode and allow you to check output in the terminal.

If running in debug mode you see an error such as:

AttributeError: 'module' objet has no attribute 'load'

try forcing a reinstall of dependencies:

pip install --force-reinstall -U -r requirements.txt

This happens when the ethernet cable is unplugged from either the task laptop, the host PC, or both.

1. Check that the cable is plugged into both computers. Try starting the experiment again.
2. Unplug and replug the cables and try again.
3. See if a zombie process is hogging the TCP port and kill by PID if so: lsof -i tcp:8889

For testing host applications, there is a debugger script included. You can generate scripted messages from the host PC's output.log (see tests/FR1_output.log). Example:

$ python -m ramcontrol.debugger.py generate -s test -x FR1 -n 1 -o fr1_script.csv -f tests/FR1_output.log

To run the generated script:

$ python -m ramcontrol.debugger.py run -f fr1_script.csv

Caveat emptor: Make sure you have loaded the experimental configuration on the host PC first!

Unit testing is in the process of being added. To run existing tests, you'll need to run

$ pip install -r requirements.txt
$ pip install -e .

Then tests are run with

$ pytest