Python-minimal-install

A minimal Python environment for reproducible research in human movement sciences.

Minimal install

The minimal requirements to start with a human movement data analysis project are :

• python : the language
  • numpy and matplotlib : the python packages for scientific data analyses
  • ipykernel: to run jupyter notebooks mixing python and markdown blocks
• VSCode : the development environment (editor, debugger...)
• conda : the package management system to manage your python environment

You will need approximately 1.5 GB of free space on your hard drive to install all the needed tools.

Install python

Important

Please follow this procedure even though you already have a working python installation. ⚠️
miniconda ensures that you do not break your system's Python installation(s): it will install a separate new Python environment that you can safely use for your projects.

Install python

• Install miniconda to get a minimal python install (https://conda.io/miniconda.html).
  • Download the Latest Miniconda Installer appropriate for your computer (Mac users: prefer the pkg version)
  • Install miniconda, accepting all defaults
• Check you installation
  • Open a shell window (shell is here a generic term for Terminal on Macs and Powershell on Windows)

    1. The shell prompt should look like: (base) userName@computerName currentDirectory %
       (e.g., (base) denismottet@MacBook-Pro-de-Denis % )
       The prefix (base) indicates that conda has activated the base environment automatically (default miniconda behavior).

    2. At the shell prompt, ask where is the python software located
       i.e., type which python (Terminal) or where python (Powershell). The answer should be :
       youHomeDirectory/miniconda3/bin/python (Terminal)
       youHomeDirectory\miniconda3\bin\python (Powershell)

• If you do not get the previous results :
  • double check the options you selected when installing miniconda
  • double check the version of the installer you have chosen

Install VSCode

VSCode is a development environment well suited for python.

• Install VSCode
  • download the appropriate version for you computer from https://code.visualstudio.com/
  • install VSCode, accepting all defaults
• Open VSCode and install the required extensions for python
  • VS Code Python extension, to edit,run and debug the python code.
  • Jupyter Extension for Visual Studio Code, to edit,run and debug the jupyter notebooks.

Install the required python packages

The goal is to install the packages in the base environment (i.e., your default python environment).
Open a shell window to run the install commands

• install numpy, to manipulate data series and matrices
  • conda install numpy
• install matplotlib, to plot data
  • conda install matplotlib
• install scipy, to perform scientific computations
  • conda install scipy
• install ipykernel and nbformat, to run jupyter notebooks in VSCode
  • conda install ipykernel nbformat
• install the matplotlib widget backend, to display interactive plots in jupyter notebooks
  • conda install -c conda-forge ipympl

Test you environment

Once you have installed all the required tools, you can test your environment by running a simple Jupyter notebook in VSCode.

• Open VSCode
• Create a new file (File -> New File)
• Save the file (File -> Save As) in a new directory (e.g., python-minimal-install-test) with the name test.ipynb
• In the file, type the following code in a cell and run it (Run -> Run Cell)

import numpy as np
import matplotlib.pyplot as plt

x = np.linspace(0, 2*np.pi, 100)
y = np.sin(x)

plt.plot(x, y)
plt.show()

• You should see a plot of the sine function in the test.ipynb file.

Next steps

• Learn how to use VSCode and Jupyter to develop and document your ideas (text + code).
• Learn how to use Python-for-HMS-Template to save time and minimize errors in your projects.
• Learn how to use Git to manage the evolution of your ideas, and GitHub to share them with others.

Bonus install

The following tools are not required to start with a human movement data analysis project, but they will be useful later on.

Create a GitHub account (for you as a student)

Github is a cloud-based platform that lets you track and control changes to your code. As a student, you get free pro access to some very useful tools.

• go to https://github.com/ and sign up
  • e-mail: your student's email
  • username: your First name + the first 3 letter of your Familly name + _student (e.g. DenisMot_student)

Install GitHub Desktop

GitHub Desktop is an easy front end to the Git version control system.

• Install GitHub Desktop
  • download the appropriate version for you computer from https://github.com/apps/desktop
  • install GitHub Desktop, accepting all defaults

Other install options

If you are interested in a more complete install, you can install anaconda instead of miniconda.
Yet, the anaconda install is a big install (4.5 GB) that comes with a lot of packages that you do not need (at the beginning).
Take a look at the anaconda vs miniconda web page https://docs.anaconda.com/distro-or-miniconda/ before making your choice.