Code for the preprint: "Behavioral characteristics of dopamine D5 receptor knockout mice: locomotor activity, working memory, learning, and impulsivity-related behaviors."

This reporsitory includes code for assessment of learning process in q-learning.

Data

• data/merged_structured_df.csv : Experiment results of 3-choice serial reaction time task.
• data/q_learning_10Aug2021.h5 : Sampling results of q-learning model.
• data/dt_q_case1 : Individual fitting results for q-learning model.

Environments

• docker-compose.yml : Type docker-compose up to launch up JupyterLab.

Docker image, toshiaki0910/3csrtt:v03 will be fetched from DockerHub. This image was built from Dockerfile.
Caution : When you perform JupyterLab operations, jupyter-vim was installed in the docker image. For detail, see https://github.com/jwkvam/jupyterlab-vim.

Code

• src/main_q_learning_sampling_results.ipynb : This notebook does sampling and produces plottings.
• src/cpy_q_learning.pyx : Core part of q-learning model and is compiled by Cython.
• src/q_learning.py, src/q_utils.py : Cleaning and helper functions to run sampling and plottings.

Figures

• dt_figs : Directory containing figures used for the manuscript.