Code to run the Red-Green Experiment Locally

Requirements:

• Desktop/Laptop
• Conda
• Git

Step 1: First clone the repo (use https cloning if no ssh key) and enter the directory

git clone git@github.com:arijit-dasgupta/red_green_experimental_site.git
cd red_green_experimental_site

Then install the python requirements in conda

conda create -n redgreen_exp python=3.11
conda activate redgreen_exp
pip install -r requirements.txt

Step 2: Install Node.js and npm (if not already installed)

My npm version is 10.9.0 and Node.js version is v23.3.0. To be safe, please install these versions

For Windows:

• Download and install Node.js from https://nodejs.org/
• This will automatically include npm

For macOS:

• Using Homebrew: brew install node
• Or download from https://nodejs.org/

For Linux (Ubuntu/Debian):

• Update your package index: sudo apt update
• Install Node.js and npm: sudo apt install nodejs npm

Step 3: Build the frontend (ignore the barrage of deprecation warnings)

npm --prefix frontend install
npm --prefix frontend run build

Step 4: Run the experiment and go to the localhost link shown in the terminal (default is http://127.0.0.1:5000)

python backend/run_redgreen_experiment.py

The data will be automatically saved in a SQLite database .db in a folder named instance from the folder where the experiment python file is run (in the example above, it will be in the root folder).

Notes on running experiment (from Arijit)

The backend/trial_data stores all the datasets. The dataset present here is named pilot_final which was used for CogSci 2025's JTAP experiments. in the dataset folder has all the familiarization stimuli and experiment trial stimuli. Fam trials start with 'F' and an integer starting from 1 until the max number of trials (So F1, F2 etc.). The same is the case for the main experiment trials (E1, E2, E3 etc.) which start with an 'E'.

This experiment creates an SQL database using flask_sqlalchemy in the backend and handles trial randomization, assignment, counterbalancing, and fine-grained keystroke-per-frame data recording. To configure the experiment, edit the start of backend/run_redgreen_experiment.py. You can monitor the experiment using backend/experiment_monitoring_dashboard.py, though stability is not guaranteed. For database inspection during the experiment, I usually open the database file in a GUI such as DB Browser for SQLite.

To postprocess the database file. Run through the backend/postprocess.ipynb notebook and configure the values in the first cell. Then run through the cells to save a .pkl file. Optionally, you can write your own postprocessing code, but I highly recommend you to re-use the extract_human_data function from backend/postprocess_redgreen_human_data.py, so that you don't have to deal with SQLAlchemy, you simply get the data in pandas dataframes. The demographics csv file is the one from Prolific. If you do not have it, you can comment out that section.

Keep in mind that the end of the experiment points back to a Prolific return link for participants in frontend/src/pages/Finish.js around line 90-100. Feel free to change this for your use case. If you are using prolific, you MUST change the code here.