.. _getting-started-label:

Getting started with diff_classifier
====================================

Pip install
-----------
Currently, diff_classifier is not on PyPi, but the next version of
diff_classifer will be uploaded to PyPi.

Git clone
---------
Users can clone a copy of diff_classifier with the command

.. code-block:: python

  git clone https://github.com/ccurtis7/diff_classifier.git

Running the setup file will install needed dependencies, including Fiji:

.. code-block:: python

  python3 setup.py develop


You can install diff_classifier from the `Github repository
<https://github.com/ccurtis7/diff_classifier>`_.  This will install
diff_classifier and its Python dependencies. Users will also need to have
downloaded `Fiji <https://imagej.net/Fiji/Downloads>`_ in order to implement the
Trackmate functionality.

Checking Fiji installation
--------------------------
The tracking portion of diff_classifier relies on the Fiji plugin `Trackmate
<https://imagej.net/TrackMate>`_. The setup.py file should do this automatically.

Otherwise, users can install Fiji manually at `here <https://fiji.sc/#download>`_.
An example installation using the shell:

.. code-block:: shell

  wget https://downloads.imagej.net/fiji/latest/fiji-linux64.zip
  unzip fiji-linux64.zip

When using the diff_classifier.imagej module, default behavior looks for Fiji at
~/Fiji.app/ImageJ-linux64.exe for Linux users or
/Applications/Fiji.app/Contents/MacOS/ImageJ-macosx.exe for Apple users.
Otherwise, users must manually specify the location using the fiji_bin variable.


Note: Python2/Python3 compatibility
-----------------------------------

Diff_classifier is not guaranteed to run on Python2. 

Examples
--------
You can find example implementations of diff_classifier in the `notebooks
directory <https://github.com/ccurtis7/diff_classifier/tree/master/notebooks>`_
on Github.