ReplicantDriveSim 0.1.5

Traffic simulation package with C++ backend

---

layout: default title: "Traffic Architecture" permalink: /External/

Build and Install the Traffic Simulation Library

Verifying Git LFS Installation

To check if Git LFS is correctly installed and initialized, you can run:

git lfs version

Check if specific files are being tracked by Git LFS using:

git lfs ls-files

Fetching LFS Files:

git lfs pull

Initialize Submodules

This will initialize and clone the submodule repository and check out the appropriate commit specified in the parent repository.

git submodule update --init --recursive

Verify the submodule status

Ensure that the submodule is now populated correctly.

git submodule status

Environment setup

conda env create -f environment.yml
conda activate drive

Build and check setup

export MACOSX_DEPLOYMENT_TARGET=14.0
TRAFFIC_SIM_SOURCEDIR=$PWD python -m build -v
unzip -l dist/simulation-*.whl

or configure and build a stand alone traffic library

mkdir build
cd build
cmake ..
make
cd ..

This will build the Google Test library (libgtest.a and libgtest_main.a).

Install Traffic Simulation

pip install --force-reinstall dist/simulation-*.whl

Example Usage

## Import the compiled C++ module
import traffic_simulation

## Create a traffic environment
simulation = traffic_simulation.TrafficSimulation(2)

## Retrieve the current states of agents in the traffic environment
states = simulation.get_agent_positions()

## Display the states of agents in the traffic environment
for agent, state in states.items():
    print(f"{agent} state: {state}")

## Advance the environment by one step
simulation.step([1, 0], [[0.1, 0.2, 0.3], [0.0, 0.0, 0.0]])

## Update the states of agents in the traffic environment
states = simulation.get_agent_positions()

## Display the updated states of agents in the traffic environment
for agent, state in states.items():
    print(f"{agent} state: {state}")

Docker build

# Build docker image
export DOCKER_BUILDKIT=1 
docker-compose up --build
# Run docker image
docker run -it --rm -v $(pwd)/data:/app/repo/data replicantdrivesim-app bash

# Alternative build method
DOCKER_BUILDKIT=1 docker build --ssh default -t replicantdrivesim-app .

# Close docker container
docker-compose down
docker-compose down --volumes
docker-compose down --volumes --remove-orphans

docker system prune -a
docker system prune -a --volumes

docker-compose ps -a
docker-compose logs

# Exec into the Running Container (app is name of service in docker-compose.yml)
docker-compose exec app bash

python simulacrum.py

# Or

python trainer.py

Monitor simulation session

mlflow ui --backend-store-uri file:mlruns

Run unit tests

./build/tests/traffic_simulation_test
./build/tests/perception_module_test

Google Test provides robust features for writing and organizing unit tests in C++. Customize your test structure (TEST_F, TEST, etc.) as per your project requirements.

Steps to Create a Git Tag and Push It to a Remote Repository

Create a New Tag Locally

To create an annotated tag, use the following command. Replace v1.0.0 with your desired tag name and customize the message as needed:

git tag -a v1.0.0 -m "Release version 1.0.0"

• -a v1.0.0: This option creates an annotated tag with the name v1.0.0.
• -m "Release version 1.0.0": This option adds a message to the tag, which is stored with it.

Push the Tag to the Remote Repository

Once the tag is created locally, push it to the remote repository:

git push origin v1.0.0

• origin: The name of the remote repository (typically origin by default).
• v1.0.0: The name of the tag you created.

Triggering the Workflow Manually via the Command Line

Trigger the Workflow Using curl:

curl -X POST \
  -H "Authorization: token GITHUB_PERSONAL_ACCESS_TOKEN" \
  -H "Accept: application/vnd.github.v3+json" \
  https://api.github.com/repos/YOUR_USERNAME/YOUR_REPOSITORY/actions/workflows/YOUR_WORKFLOW_FILE.yml/dispatches \
  -d '{"ref":"main"}'

• GITHUB_PERSONAL_ACCESS_TOKEN: Your personal access token.
• YOUR_USERNAME: Your GitHub username.
• YOUR_REPOSITORY: Your repository name.
• YOUR_WORKFLOW_FILE.yml: The filename of your workflow YAML (e.g., publish.yml).
• main: The branch you want the workflow to run against.

Pulling Docker Images from DockerHub Registry

This section provides instructions on how to interact with DockerHub to pull Docker images, including steps for accessing private repositories and running containers with shell access.

DockerHub Registry

Docker Login

If the Docker image repository is private, you will need to authenticate with Docker before pulling the image. To log in, use the following command:

docker login

Enter your Docker Hub credentials when prompted. After successfully logging in, you will have access to pull and run the private repository images.

Running the Container with Port Mapping and Shell Access

To run the Docker container with port mapping and gain interactive shell access, use the following command:

docker run -it -p 8080:80 chrisjcc/replicantdrivesim /bin/bash

This command will map port 8080 on your host to port 80 in the container and start an interactive Bash shell within the container.

Running the Container with Shell Access

If you only need shell access without port mapping, you can run the container interactively using:

docker run -it chrisjcc/replicantdrivesim /bin/bash

This command starts the container and opens a Bash shell, allowing you to interact directly with the container environment.

Create a Release

Follow the instructions below to generate a Git tag, which will trigger a software release.

git fetch --tags  # Retrieve all tags from the remote repository
git tag -a v0.1.0 -m "Version 0.1.0"  # Create a new annotated tag for version 0.1.0
git push origin v0.1.0  # Push the newly created tag to the remote repository
git fetch --tags  # Refresh the local tags list
git tag -l  # List all tags to verify the new tag is present