Installation
=================
.. note::

   **Recommended Install**
    CRISPy is actively being developed at the moment,
    install from the source (see Option 3) to stay up-to-date with the latest version.

System Requirements
--------------------
CRISPy is compatible with the following environments:
- Python 3.8 or later.
- Dependencies include `numpy`, `astropy`, `scipy`, `scikit-image`, and `scikit-learn`.

Instructions
-------------

1. **Install from PyPI**:

   The easiest way to install CRISPy is through PyPI. Run the following command in your terminal:

   .. code-block:: bash

       pip install crispy-learn

2. **Install from Source**:
   to install the latest developing version, clone the CRISPy GitHub repository:

   .. code-block:: bash

        git clone https://github.com/mcyc/crispy.git
        cd crispy
        pip install -e .


   To use a specific version using a tag, for example, v1.2.1, run the following after
   the initial install:

   .. code-block:: bash

        git checkout v1.2.1
        git pull

3. **Dependencies**:
   CRISPy will automatically install its dependencies during installation.

   If you encounter issues with pre-existing versions of dependencies or want to ensure
   that the pinned versions of CRISPy's dependencies are installed, use the following command:

   .. code-block:: bash

       pip install --upgrade --force-reinstall --no-cache-dir -e .

   The above command ensures that all dependencies are freshly installed, replacing any older or conflicting versions.
   Alternatively, if any issues occur, install the dependencies manually:

   .. code-block:: bash

       pip install numpy scipy astropy joblib scikit-image scikit-learn

Verification
~~~~~~~~~~~~~
To verify the installation, open a Python interpreter and import CRISPy:

.. code-block:: python

    import crispy
    print(crispy.__version__)

If no errors occur, CRISPy is installed correctly.