Installation

Prerequisites

  • Python 3.8 or higher

  • Pip (Python package manager)

Dependencies

Le packge nécessite les bibliothèques suivantes :

  • numpy >= 1.20.0

  • scipy >= 1.7.0

  • cvxpy >= 1.2.0

  • matplotlib >= 3.5.0

  • pandas >= 1.3.0

  • tkinter (included with Python)

Installation from PyPI

pip install BsplineQuantRegpy

Installation from source

git clone https://github.com/username/BsplineQuantRegpy.git
cd BsplineQuantRegpy
pip install -e .

Installing dependencies

pip install -r requirements.txt

Verifying the installation

import BsplineQuantRegpy
print(BsplineQuantRegpy.__version__)

Supported SOCP solvers

The package supports several solvers for SOCP optimization:

  • CLARABEL: Default solver, open-source

  • ECOS: Open-source solver, fast for small problems

  • SCS: Open-source solver, suitable for large problems

  • MOSEK: Commercial solver, high-performance (requires a license)

  • GUROBI: Does not support SOCP management. Only usable for splines of degree 1 or 2.