Welcome to Codpy, the book !

This site is the companion site to the book Reproducing Kernel Methods for Machine Learning, PDEs, and Statistics with Python, available on arXiv by Philippe G. LeFloch, Jean-Marc Mercier, and Shohruh Miryusupov.

The main purpose of this companion site is to provide facilities to reproduce all figures of the book, in order to ease the diffusion of RKHS methods to teachers, students and practitioners. It allows to

• facilitate the understanding of the material presented in the book, and to provide a hands-on introduction to CodPy, an open-source Python library for numerical algorithms, machine learning, and computational statistics.

• enable readers to experiment with the algorithms presented in the book, and to modify the code for their own purposes.

• offer a platform for sharing additional resources, such as datasets, code snippets, and tutorials related to the book’s content.

About CodPy

CodPy (Curse Of Dimensionality in Python) is an open-source, RKHS dedicated, Python library designed for numerical algorithms, machine learning, and computational statistics. It has a wide range of applications in finance, engineering, and industry, and its technical documentation is located at this url.

Installation

This has been tested on Windows 11.

1. Make a python 3.9 virtual environment

Note: Python 3.9 is required

python --version
python 3.9.13

Create a virtual environment:

python -m venv venv
venv\Scripts\activate

2. Clone the github repository:

git clone https://https://github.com/JohnLeM/codpybook-rtd.git

3. Install the dependencies

cd codpybook
pip install -e .

The python files can be found at: codpybook/docs/chX

You can already run the files to reproduce the experiments available at Readthedocs

4. Compiling with sphinx

If you want to re-compile all the experiments, as you see them on the website, you can:

cd docs
sphinx-build html -b . _build

Once done, you can find the index.html file in: codpybook/docs/_build/index.html

Alternatively, you can also install the package directly from GitHub:

pip install git+https://github.com/JohnLeM/codpybook-rtd.git

Report Issues

Please report bug tracking and enhancement requests to our GitHub page

Chapter 2:

• Chapter 2: Basic notions on reproducing kernels

Chapter 3:

• Chapter 3: Discrete operators based on reproducing kernels

Chapter 4:

• Chapter 4: Clustering strategies

Chapter 5:

• Chapter 5: Optimal transport and statistical kernel methods

Chapter 6:

• Chapter 6: Applications to machine learning : supervised, unsupervised and generative methods

Chapter 7:

• Chapter 7: Application to partial differential equations

Chapter 8:

• KQLearning
• Chapter 8: Reinforcement learning.

Chapter 9:

• Chapter 9: Mathematical Finance.

Indices and Tables

• Index

• Module Index

• Search Page