State control system for automated, high-throughput behavioral training based on Python. It is tightly intergated with Database storage & control using the Datajoint framework. It can run on Linux, MacOS, Windows and it is optimized for use with Raspberry pi boards.

It is comprised of several overridable modules that define the structure of experiment, stimulus and behavioral control. A diagram that illustrates the relationship between the core modules:

Core modules:

Main state experiment Empty class that is overriden by other classes depending on the type of experiment.

This class can have various State classes. An Entry and Exit State are necessary, all the rest can be customized.

A typical experiment state diagram:

Each of the states is discribed by 4 overridable funcions:

Tables that are needed for the experiment that discribe the setup:

SetupConfiguration
SetupConfiguration.Port
SetupConfiguration.Screen

The experiment parameters are specified in *.py script configuration files that are entered in the Task table within the lab_experriment schema. Some examples are in the conf folder but any folder that is accessible to the system can be used. Each protocol has a unique task_idx identifier that is used uppon running.

Implemented experiment types:

• MatchToSample: Experiment with Cue/Delay/Response periods
• MatchPort: Stimulus matched to ports
• Navigate: Navigation experiment
• Passive: Passive stimulus presentation experiment
• FreeWater: Free water delivery experiment
• Calibrate: Port Calibration of water amount
• PortTest: Testing port for water delivery

Empty class that handles the animal behavior in the experiment.

IMPORTANT: Liquid calibration protocol needs to be run frequently for accurate liquid delivery

Implemented Behavior types:

• MultiPort: Default RP setup with lick, liquid delivery and proximity port
• VRBall (beta): Ball for 2D environments with single lick/liquid delivery port
• Touch (beta): Touchscreen interface

Empty class that handles the stimuli used in the experiment.

Implemented stimulus types:

• Grating: Orientation gratings
• Bar: Moving bar for retinotopic mapping
• Movies: Movie presentation
• Olfactory: Odor persentation
• Panda: Object presentation
• VROdors: Virtual environment with odors
• SmellyObjects: Odor-Visual objects

Non-overridable classes:

Handles all database interactions and it is shared across Experiment/Behavior/Stimulus classes non-overridable

Data are storred in tables within 3 different schemata that are automatically created:

lab_experiments:

lab_behavior:

lab_stimuli:

Handles all communication with hardware

Can be run either as a service that is controled by the Control table

sudo python3 run.py

or can specify a task_idx to run directly. After it completes, the process ends.

sudo python3 run.py 1 

This process can be automated by either a bash script that runs on startup or through control from a salt server.

Get latest raspbian OS in raspi-config:

• enable ssh
• disable screen blanking
• enable Desktop auto-login

Change hostname - Optional, but it will make it easier to identify later

sed -r -i s/raspberrypi/<<HOSTNAME>>/g /etc/hostname /etc/hostname
sed -r -i s/raspberrypi/<<HOSTNAME>>/g /etc/hosts /etc/hosts

Change username - Optional

sudo useradd -s /bin/bash -d /home/<<USERNAME>>/ -m -G sudo <<USERNAME>>
sudo passwd <<USERNAME>>
mkhomedir_helper <<USERNAME>>
sudo userdel -r -f pi

Install salt for remote control, you need to have a salt-master server! - Optional

sudo apt install salt-minion -y
echo 'master: <<YOUR_SALT-MASTER_IP>>' | sudo tee -a /etc/salt/minion
echo 'id: <<HOSTNAME>>' | sudo tee -a /etc/salt/minion
echo 'master_finger: <<MASTER-FINGER>>' | sudo tee -a /etc/salt/minion
sudo service salt-minion restart

X display settings for ssh run, important for Panda stimulus

sed -i -e '$aexport DISPLAY=:0' ~/.profile
sed -i -e '$axhost +  > /dev/null' ~/.profile

Install dependent libraries

sudo apt update
sudo apt install python-dev libatlas-base-dev build-essential libavformat-dev libavcodec-dev libswscale-dev libsquish-dev libeigen3-dev libopenal-dev libfreetype6-dev zlib1g-dev libx11-dev libjpeg-dev libvorbis-dev libogg-dev libassimp-dev libode-dev libssl-dev libgles2 libgles1 libegl1 -y

Install python packages

sudo pip3 install 'numpy>=1.19.1' pygame==1.9.6 cython pybind11 scipy datajoint omxplayer-wrapper imageio imageio-ffmpeg

Install correct multitouch driver for 7" inch raspberry pi screen

git clone http://github.com/ef-lab/python-multitouch ~/github/python-multitouch
cd ~/github/python-multitouch/library
sudo python3 setup.py install

Install panda3d version for raspberry pi

wget ftp://eflab.org/shared/panda3d1.11_1.11.0_armhf.deb
sudo dpkg -i panda3d1.11_1.11.0_armhf.deb

Enable pigpio service

wget https://raw.githubusercontent.com/joan2937/pigpio/master/util/pigpiod.servicesudo cp pigpiod.service /etc/systemd/system
sudo systemctl enable pigpiod.service
sudo systemctl start pigpiod.service

Get PyMouse

git clone http://github.com/ef-lab/PyMouse ~/github/PyMouse

Create dj_local_conf.json with the correct parameters in the PyMouse folder:

{
   "database.host": "YOUR DATABASE",
    "database.user": "USERNAME",
    "database.password": "PASSWORD",
    "database.port": "PORT",
    "database.reconnect": true,
    "database.enable_python_native_blobs": true
}

Create tables

cd ~/github/PyMouse
python3 -c 'from core.Experiment import *'
python3 -c 'from core.Stimulus import *'
python3 -c 'from core.Behavior import *'
python3 -c 'from Stimuli import *'
python3 -c 'from Behaviors import *'
python3 -c 'from Experiments import *'