Advanced Python for science

This is a website for a training on Python for science written by the group py-edu-fr.

Content

• About this training
• Preliminaries
  • About setting up a computer for scientific computing
  • Notes for Windows users
  • Notes for macOS users
  • Notes for Linux and WSL users
  • Python
  • Conda ecosystem
  • Editors/IDE
  • Versioning tools
  • Clone the repo of this training
• Appetizer with JupyterLite
  • First steps pure Python: function calls
  • First steps pure Python: objects, types and variables
  • First usage of Python libraries
• Generalities on Python
  • Get and use the repository of this training
  • Main characteristics
  • Python limitations
  • Different ways to use Python
  • Tools to develop Python code
• Basic Pure Python
  • Simple types, variables
  • Strings, bytes and immutability
  • Conditions (if and match)
  • Lists and tuples
  • Loops (while and for)
  • Exceptions and traceback
  • Read / write files
  • More data structures
• Structure and reuse code
  • Functions (very basic)
  • Testing
  • Practice: Read corrupted files
  • Modules, packages and imports
  • Classes and objects
  • The standard library
  • Documentation
  • Back to functions (less basics)
• Use, make & share Python software
  • Introduction: PyPI and friends
  • Conda & the conda-forge project
  • Global installation of Python apps
  • Using Third-party libraries
  • Modern Python packaging
• Scientific Python (basics)
  • Python scientific ecosystem
  • Introduction to Numpy and the Python Array API standard
  • NumPy efficiency
  • Creation of Numpy arrays
  • Manipulating NumPy arrays
  • Introduction to Matplotlib and alternatives
  • Alternatives
  • Introduction to Pandas and alternatives
• Recommended file formats
• Software engineering
  • Versioning
  • Continuous integration
  • Documentation
  • Debugging
  • Gradual typing
• Intermediate pure Python
  • Decorators
  • OOP and magic methods
  • Generators
  • OOP and heritage
  • Define context managers
• Performance optimization
  • Generalities on optimization in sequential
  • Profiling
• Other languages and wrapping
  • Wrapping C and C++
  • Wrapping Rust
  • Wrapping Fortran
• Parallelism
  • Threading
  • Concurrency
  • Multiprocessing
  • MPI and mpi4py
• Tools for Python HPC
  • Cython
  • Pythran
  • Numba
  • JAX
  • Codon
• Application on numerical kernels
  • NBody
  • DTW
• A bit of Web
• Native Python applications
• GPU
• Databases
• Future & science fiction
  • Subinterpreters
  • Free Threading Python
  • Faster CPython and CPython JIT
  • HPy
  • Todo list

Todo

Quick intro to file format

(The original entry is located in /builds/py-edu-fr/py-edu-fr/books/science/book/file-formats/hdf5-netcdf.md, line 16.)

Todo

• Editors / IDE (JupyterLab, VSCode/VSCodium and Spyder)

  • basis

  • main extensions for Python

  • debug

  • sync configuration

  • overall comparison between these IDEs

(The original entry is located in /builds/py-edu-fr/py-edu-fr/books/science/book/generalities/dev-tools.md, line 78.)

Todo

Check before December 2025

• First intro immutability

• First sequence: indexing, slicing, lookup (keyword in)

• Notions of objects and methods

• str (unicode)

• bytes (utf-8, latin1, …)

(The original entry is located in /builds/py-edu-fr/py-edu-fr/books/science/book/pure-python-basic/strings.md, line 271.)

Todo

Add more explanations on lock files and reproducibility.

(The original entry is located in /builds/py-edu-fr/py-edu-fr/books/science/book/py-software/libraries.md, line 110.)

Todo

Better explain different aspects of packaging (with testing, linting, formatting and doc)

We base the presentation on a simple tool to visualize info from Wikipedia using packages like sys, os, pathlib, argparse and third-party dependencies like Requests and Textual…

Documentation with Sphinx + mention MkDocs and Readthedocs

(The original entry is located in /builds/py-edu-fr/py-edu-fr/books/science/book/py-software/packaging.md, line 132.)

Todo

• Based on https://matplotlib.org/, in particular https://matplotlib.org/stable/users/explain/quick_start.html and https://matplotlib.org/stable/gallery/index.html.

• Mention Vega-Altair

(The original entry is located in /builds/py-edu-fr/py-edu-fr/books/science/book/scientific/matplotlib.md, line 16.)

Todo

• Based on Numpy documentation (https://numpy.org/), in particular https://numpy.org/doc/stable/user/absolute_beginners.html.

• Mention the Python Array API standard

(The original entry is located in /builds/py-edu-fr/py-edu-fr/books/science/book/scientific/numpy.md, line 16.)

Todo

• Based on https://pandas.pydata.org/, in particular https://pandas.pydata.org/docs/user_guide/10min.html and https://pandas.pydata.org/docs/getting_started/intro_tutorials/index.html

• Also mention [Polar], in particular https://docs.pola.rs/user-guide/getting-started/

(The original entry is located in /builds/py-edu-fr/py-edu-fr/books/science/book/scientific/pandas.md, line 16.)

Todo

Check standard library

• Small examples from (or just mention) sys, os/shutils, copy, pathlib/glob argparse, math, re, dataclasses, itertools, turtle, functools, unittest, collections, datetime

• https://docs.python.org/3/tutorial/stdlib.html

• Mention good alternatives outside of the standard library (pytest, click, requests/urllib3, typing-extensions, python-dateutil/pendulum, fsspec, certifi, idna, …)

(The original entry is located in /builds/py-edu-fr/py-edu-fr/books/science/book/structure-reuse-code/standard-library.md, line 222.)