getting_started.md

Getting started

Making a clean python environment

When starting a new python project it is useful to create a new python environment to avoid any conflicting dependencies between projects. Depending on how python was installed on your system this can be done in a few different ways.

Using venv

Python comes with its own environment manager called venv. To make a new environment named discnet_env run:

python3.10 -m venv discnet_env

This will make a new folder called discnet_env in the directory where you run the command. This directory contains the python environment.

This will make an environment using the same version of python used to run the command.  By using `python3.10` we ensure it will be a python 3.10 environment.

To activate the environment run (from the same directory as the previous command):

source discnet_env/bin/activate

Using conda

If you installed python using conda you should use it as your environment manager. To make a new environment named discnet_env run:

conda create --name discnet_env python=3.10

This will make a new directory called discnet_env inside your conda install's envs folder (typically in your home folder). This directory contains the python environment.

Unlike `venv` you can specify any version of python when making an environment even if that version is not currently installed on your system.

To activate the environment run:

conda activate discnet_env

Clone the repository

Use git to clone the repository for this workshop:

git clone https://github.com/CKrawczyk/DISCnet_workshop.git
cd DISCnet_workshop

Installing the workshop's python package

Throughout this workshop you will be helping to write a python package called data_transforms. To install it into your new environment (once the env is activated) run:

# make sure `pip` is up to date and install the `flit` build tool
pip install -U pip flit

# install the development version of the `data_transforms` package
pip install -e .[dev]

The option -e installs the code in "edit" mode, this means the package directory is sym-linked into your python path. Any changes you make to the code will automatically be "installed" without needing to run the pip command again. .[dev] indicates you want to install the python package located in the current folder and to also install the optional development dependencies for the package. We will cover these topics in more detail in the "packaging python code" section of the workshop.

Running tests

The notes on test driven development will go into this in more detail, but for completeness you can run this codes tests with the command:

coverage run

and view the test coverage report with:

coverage report