Libraries

Imports

Libraries are included with the keyword import.

import math

math.cos(2)

import collections as col  # Change the namespace with `as`

counter = col.Counter(["a", "b", "c"])

Common built-in libraries.
Library	Use
`math`	Mathematical functions
`os`	Operating system interfaces
`re`	Regular expressions
`sys`	System-specific parameters and functions

Installing

PIP installs python libraries.

pip install numpy
# Or
python -m pip install numpy

Python packages are distributed as wheels. PyPI lists almost every public wheel. PIP grabs libraries from here.

Common external.
Library	Use
`flask`	Web design
`matplotlib`	Plotting
`numpy`	Matrices and arrays
`pandas`	Data tables
`scipy`	Scientific computing
`torch`	Neural networks

Environments

For a project, it is best practice to use a virtual environment instead of installing libraries globally. This is done with the python module venv.

python -m venv .venv  # Make the directory .venv to contain the libraries
source .venv/bin/activate  # Activate environment on Linux or MacOS
.venv/Scripts/activate  # Activate environment on Windows
deactivate  # Leaves environment

Furthermore, dependencies should be listed in a file that is named requirements.txt with their version.

Contents of requirements.txt. `

numpy==1.2
pandas>=2.1

pip install -r requirements.txt  # Install requiremements