random_matrices

Generators for several types of random matrices in Python.

Important note: This repository has just recently been made public and is still undergoing debugging and development.

Available ensembles of random matrices

• GOE: Gaussian Orthogonal Ensemble
• SG_0_+: positive-definite matrices with the identity matrix as mean value
• SG_eps_+: positive-definite matrices with identity matrix as mean value and with a positive lower bound
• SE_0_+: positive-definite matrices with given mean value
• SE_+0: positive semidefinite matrices with given mean value
• SE_rect: rectangular matrices with given mean value

Installation

First, clone the repository in the folder of your choice using:

git clone [email protected]:rcapillon/random_matrices.git

Then, activate the virtual environment for your project and install this package and its requirements with:

cd random_matrices/
pip install .

Usage

Choose and import the desired generators from generators.py, e.g.:

from random_matrices.generators import SE_0_plus

Users are invited to look up the necessary arguments to generate samples of a random matrix.

Bibliography

[1] C. Soize, Uncertainty Quantification, 2017, https://doi.org/10.1007/978-3-319-54339-0