.. _getting_started_label:

====================================
Getting Started with Reproducible-ML
====================================

We recommend to setup Miniconda_ or Conda_ to create a virtual environment.

.. _Miniconda: https://conda.io/miniconda.html
.. _Conda: https://conda.io/docs/

Creating a virtual environment
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

1. Create a virtual environment from the environment file, which can be found in the home directory of this project:

.. code-block:: console

   conda env create -f environment reproducible

2. Activate the new environment:

.. code-block:: console

   source activate reproducible

Set up MongoDB
^^^^^^^^^^^^^^
We recommend MongoDB since it is a noSQL database and it allows easy storage of arbitrary JSON
documents. Sacred :cite:`Sacred` itself recommends the MongoDB.

Start by downloading the binaries for Ubuntu 16.04:

.. code-block:: console

   mkdir mongodb
   cd mongodb
   wget https://fastdl.mongodb.org/linux/mongodb-linux-x86_64-ubuntu1604-4.0.3.tgz


Add path to your .bashrc file:

.. code-block:: console

   export PATH="$HOME/mongodb/bin:$PATH"

Then you can start the MongoDB daemon:

.. code-block:: console

   mkdir reproducible-db
   mongodb --dbpath reproducible-db --fork --logpath reproducible-db

More information concerning the MongoDB can be found here_.

.. _here: https://docs.mongodb.com/

How to use Sacredboard
^^^^^^^^^^^^^^^^^^^^^^
Having set up the Mongo database, you can track your experiments in the database by just adding -m sacred:

.. code-block:: console

   python <train.py with params> -m sacred

This will create a new entry in the Sacred MongoDB. If you access a server via ssh make sure you open a port. e.g.:

.. code-block:: console

   ssh username@someserver.ch -L 10000:localhost:10000

Then you can open the Sacredboard via:

.. code-block:: console

   sacredboard -m sacred --port 10000 --no-browser

Workflow
^^^^^^^^

Since all is ste up now, you can try to run some experiments and tests.

From the root folder :file:`/reproducible-ml` you can run experiments as modules:

.. code-block:: console

   python -m exps.brain.gan

Preferably you run your Code on a GPU. If you want to run it on CPUs change the tensorflow-gpu package
to the TensorFlow package in your environment.

There are several tests you can run:

.. code-block:: console

   python -m unittest datasets.brain.test_serialize

Or run all tests at once:

.. code-block:: console

   python -m unittest