Simulation Methods for the Circular Drift Diffusion Model (CDDM)

This repository contains R code for the implementation of different simulation methods for generating data from the Circular Drift-Diffusion Model (CDDM). The performance of each method is evaluated for both computational efficiency and statistical accuracy.

Overview

File cddm_fullReport.Rmd contains a brief description of the simulation methods explored, and the custom functions build for their implementation. The four simulation methods explored are:

1. Direct Emulation: Approximates the underlying stochastic process through discrete time steps
2. Rejection Sampling: Operates in a bounded bivariate space
3. Metropolis-Hastings Algorithm: Uses a bivariate normal proposal distribution
4. Probability Integral Transform: Applies probability integral transform using trapezoid integration

Each method generates paired angular choices (in radians) and response times (in seconds). They are evaluated on the basis of:

• Their execution time
• The difference between the mean angular choice generated and the true underlying drift direction used to generate the data.
• The difference between the empirical cumulative distribution function of the generated data and the theoretical cumulative distribution function approximated through numerical integration.

Repository Structure

• code/cddm: R code specific to the CDDM
• code/ddm/: R code specific to the DDM
• code/figures: R scripts for generating figures
• code/general_functions: R functions for general use
• code/tests: R scripts for testing the simulation methods
• references/: References and citations
• tests/: Placeholder for analysis scripts

Docker Container for Reproducibility

To ensure reproducibility, we provide a Docker container with all necessary R packages and dependencies.

Getting Started

Prerequisites

• Install Docker on your system

Building and Running the Container

1. Clone this repository:

git clone git@github.com:Adrifelcha/cddm-simulationMethods.git

2. Move to the code directory:

cd cddm-simulation-methods

3. Build the Docker image:

docker build -t cddm-simulation-methods .

3. Run the container:

docker run -p 2222:22 cddm-simulation-methods

This will start the container and expose the SSH service on port 2222.

4. Connect to the container via SSH:

ssh -p 2222 root@localhost

When prompted, write down the root password: adri93

Working with the Container

Once connected to the container:

• All required R packages are pre-installed
• The code is available in the /app/code/cddm directory
• You can run R scripts and interact with the R environment