Installation

qmat is a lightweight package to generate time-integration coefficients, which can be use as core dependency in any solver for Initial Value Problems (IVP). It depends only on numpy & scipy, and can be installed from pypi using :

pip install qmat

qmint is an prototyping package implementing different kind of IVP solvers, that can be used for learning or experiments. It has additional dependencies compared to qmat (e.g matplotlib, etc …) and can be installed from pypi using :

pip install qmint

📣 Installing qmint will automatically install qmat if it’s not already installed on your Python environment.

Install using conda

Currently, no version is distributed on conda-forge. However using pip from conda will install qmat/qmint in your conda environment.

If you are using a environment.yml file with conda, then you can add it as a dependency like this \(:\)

name: yourEnv
channels:
  - conda-forge
  - defaults
dependencies:
  ...
  - pip
  - pip:
    - qmat # or qmint

Install qmint from source

In case you want the latest revision (or a specific branch), you can directly clone the sources from github :

git clone https://github.com/Parallel-in-Time/qmint.git

If you want to use the package only, simply use the pip local installer directly :

cd qmint     # go into the local git repo
pip install .

For developers who want to contribute, recommended approach is to install the package in editable mode :

cd qmint     # go into the local git repo (if not already there)
pip install -e .[test]    # on MAC-OS : pip install -e ".[test]"

This will link your python installation to your local qmint folder, hence all your modifications will be taken into account at each new import of qmint.

🔔 Some IDEs also modify the PYTHONPATH to include the qmint root folder, which you can also do manually if you prefer.