tudelft-pycar 0.0.6.dev0

Support library for the 'EPO-4: "KITT" autonomous driving challenge' at Delft University of Technology

EPO4 Python

This repository contains Python abstractions, serial, and audio handlers for EPO4.

Documentation

The pycar documentation is generated using Sphinx v6.1.3. To open the documentation navigate to docs/build and open index.html; index.html is the documentation's start page.

Python Environment

Python environments / virtual environments are used for project isolation and dependency management. They can be set up as to specify a specific Python version to the project, and install packages locally to the virtual environment directory. Consequently, the dependencies (or packages) are isolated from the global environment, avoiding conflicts between dependency requirements of different project and allowing for the installation of a specific dependency's version.

There exist several Python tools performing all or part of these function, most notably virtualenv, pyenv, and pipenv; virtualenv allows of packages in a virtual environment, pyenv allows for several Python versions to be installed and run, pipenv is a merge between the preceding two tools. More introductory information is found on Flimm's thread titled What is the difference between venv, pyvenv, pyenv, virtualenv, virtualenvwrapper, pipenv, etc?. It is highly recommended becoming familiar with some of these tools as they are highly used in academic and private settings.

In the EPO4 project, the pipenv tool is used. When an environment is created, two files named Pipfile and Pipfile.lock are also created in the project's directory. A third file named requirements.txt can be used to create an environment, however, Pipfile always precedes the requirements file and will be used in the creation of an environment if available in the target directory. Pipfile has a [[source]] section specifying the repository containing the Python packages, a [packages] section specifying the packages and their respective version requirements, , a [dev-packages] section specifying packages for development (not installed on the machines of package users), and a [requires] section specifying Python's version. Pipfile.lock contains information regarding the installed packages and package dependencies. The environment is created in a directory outside the project. For more information on pipenv read Pipenv: Python Dev Workflow for Humans and Basic Usage of Pipenv.

Pipenv Setup (Windows)

First, open a command prompt and make sure pip is installed.

>pip --version

Note that if pip is not recognized, it is possible that pip3 is installed (with Python 3); run the same command but use pip3 instead. In case pip3 is also not recognized then install it.

Next, install pipenv (note that if pip3 is installed it should be used instead of pip).

>pip install --user pipenv

Navigate to the repository's directory and install environment.

>pipenv install

After completing these steps, the project's python environment should be installed with the specified Python version and required packages.

To use the environment (run scripts / install additional packages), it should be first activated using the pipenv shell command.

>pipenv shell

A string containing (epo4-python-'hash') is appended before the directory. Scripts should be run from inside the environment.

Pipenv Setup (MacOS)

TODO: Write instructions for MacOS.

Pipenv Setup (Linux)

TODO: Write instructions for Linux.

PyCharm Setup

To run scripts directly from PyCharm (and for additional functionality). The environment's interpreter can be chosen as the default interpreter for the project. Open a command prompt and navigate to the project's directory. Use the pipenv command to get the installation directory of the environment.

>pipenv --venv

Copy the directory's location, and in PyCharm go to File - Settings - Project: epo4-python - Python Interpreter; then go to the drop-down menu, Show All, go to the plus sign, Pipenv Environment, at the Base Interpreter select the navigation option, go to the directory you copied earlier, enter Scripts, select python.exe, click OK, and deselect Install packages from Pipfile; save the settings.

Package Dependencies

The EPO4 project uses Python 3.10; the following packages are specified in the Pipfile.

Package	Version
pyqt5	5.15.7
pyqt5-stubs	5.15.6
matplotlib
numpy
pyaudio	0.2.12
pyserial	3.5
sphinx

EPO4 Python Abstractions

Serial Port Testing

$ socat -d -d pty,raw,echo=0,b115200 pty,raw,echo=0,b115200
$ cat /dev/pts/3

Use SerialHandle to write to /dev/pts/2, the data should be written in /dev/pts/3. In case of a permission error, give permission to all users.

$ sudo chmod 666 /dev/pts/2

$ sudo apt install libxcb-xinerama0

TODO: implement full test with command interactions