.. _sphx_glr_auto_examples_pong:

NEST-pong
=========
This program simultaneously trains two networks of spiking neurons to play
the classic game of Pong.

Requirements
------------
- NEST 3.3 or later
- NumPy
- Matplotlib

Instructions
------------
To start training between two networks with R-STDP plasticity, run
the ``run_simulations.py`` script. By default, one of the networks will
be stimulated with Gaussian white noise, showing that this is necessary
for learning under this paradigm. In addition to R-STDP, a learning rule
based on the ``stdp_dopamine_synapse`` and temporal difference learning
is implemented, see ``networks.py`` for details.

The learning progress and resulting game can be visualized with the
``generate_gif.py`` script; this requires the ``imageio`` package.



.. raw:: html

    <div class="sphx-glr-thumbnails">


.. raw:: html

    <div class="sphx-glr-thumbcontainer" tooltip="See Also --------- Original implementation">

.. only:: html

  .. image:: /auto_examples/pong/images/thumb/sphx_glr_pong_thumb.png
    :alt:

  :ref:`sphx_glr_auto_examples_pong_pong.py`

.. raw:: html

      <div class="sphx-glr-thumbnail-title">Classes for running simulations of the classic game Pong</div>
    </div>


.. raw:: html

    <div class="sphx-glr-thumbcontainer" tooltip="The script is centered around a simulation of the classic game Pong in which the vertical posit...">

.. only:: html

  .. image:: /auto_examples/pong/images/thumb/sphx_glr_run_simulations_thumb.png
    :alt:

  :ref:`sphx_glr_auto_examples_pong_run_simulations.py`

.. raw:: html

      <div class="sphx-glr-thumbnail-title">Application to train networks to play pong against each other</div>
    </div>


.. raw:: html

    <div class="sphx-glr-thumbcontainer" tooltip=":Authors: J Gille, T Wunderlich, Electronic Vision(s)">

.. only:: html

  .. image:: /auto_examples/pong/images/thumb/sphx_glr_generate_gif_thumb.png
    :alt:

  :ref:`sphx_glr_auto_examples_pong_generate_gif.py`

.. raw:: html

      <div class="sphx-glr-thumbnail-title">Script to visualize a simulated Pong game.</div>
    </div>


.. raw:: html

    <div class="sphx-glr-thumbcontainer" tooltip="PongNetDopa uses the actor-critic model described in [2]_ to determine the amount of reward to ...">

.. only:: html

  .. image:: /auto_examples/pong/images/thumb/sphx_glr_networks_thumb.png
    :alt:

  :ref:`sphx_glr_auto_examples_pong_networks.py`

.. raw:: html

      <div class="sphx-glr-thumbnail-title">Classes to encapsulate the neuronal networks.</div>
    </div>


.. raw:: html

    </div>


.. toctree::
   :hidden:

   /auto_examples/pong/pong
   /auto_examples/pong/run_simulations
   /auto_examples/pong/generate_gif
   /auto_examples/pong/networks