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

Conda environment (recommended)
--------------------------------

RepLikCompare depends on the usual scientific Python stack (pandas,
numpy, scipy, scikit-learn, seaborn, matplotlib) plus two clustering /
dimensionality-reduction extras, ``hdbscan`` and ``umap-learn``. An
``environment.yml`` file is provided at the root of the repository to
create a conda environment with all of them in one go:

.. code-block:: bash

   conda env create -f environment.yml
   conda activate RepLikCompare

This is the easiest way to get a working environment, since ``hdbscan``
in particular is a compiled extension that installs more reliably from
conda-forge than from PyPI on some platforms.

Installing the package
-----------------------

Once your environment is ready (conda, or any Python environment with
``pip``), you have two options:

**Option A -- without cloning (once published on PyPI)**

.. code-block:: bash

   pip install RepLikCompare

**Option B -- from source (works today)**

.. code-block:: bash

   git clone https://github.com/regueialaa/RepLikCompare.git
   cd RepLikCompare
   pip install .

Either way installs RepLikCompare together with all of its
dependencies -- including ``hdbscan`` and ``umap-learn`` -- so it is
available in both the terminal and Jupyter notebooks.