Installation
============

This guide shows you how to build and install `mcmpy`.
There are two supported installation methods for this project:

1. Installation using Docker
2. Local installation using CMake


Installation using Docker
-------------------------

First `install Docker <https://docs.docker.com/get-started/get-docker/>`_.
Then clone the GitHub repository and build the Docker image.

.. code-block:: bash

   git clone --recursive git@github.com:DM-Lab-UvA/MinCompSpin.git
   cd MinCompSpin
   docker build -t mincompspin .

After building the image, you can run the container:

.. code-block:: bash

   docker run -p 8888:8888 -it mincompspin

To run a Jupyter notebook in the Docker container

.. code-block:: bash

    jupyter notebook --ip=0.0.0.0 --no-browser --allow-root

This returns a link beginning with ``127.0.0.1``, which can be opened in a local browser.


Local installation using CMake
------------------------------

This method requires some system dependencies. Make sure the following are installed:

- C++ compiler with C++ 11 support
- CMake >= 3.26.4
- Python >= 3.11
- NumPy

Clone the GitHub repository and run the installation script:

.. code-block:: bash

   git clone --recursive git@github.com:DM-Lab-UvA/MinCompSpin.git
   cd MinCompSpin
   bash installation.sh

.. note::

   By default, the tests are built but not run.
   To build the project without the tests, execute the following commands instead: 

.. code-block:: bash

   git clone --recursive git@github.com:DM-Lab-UvA/MinCompSpin.git
   cd MinCompSpin
   mkdir -p build && cd build
   cmake .. -DBUILD_TESTS=OFF
   make
   cd mcmpy
   pip install .

Testing
-------

After installation, the test suite can be run to verify if the installation was succesful:

.. code-block:: bash

   cd build
   ctest --output-on-failure

Using MCMPy
===========

Once `mcmpy` is installed, it can be imported in Python.

.. code-block:: python

   import mcmpy
   import numpy as np