.. _sphx_glr_auto_examples_brette_et_al_2007:

Brette et al. 2007 Benchmarks
==============================

This directory contains Python implementations of the benchmarks from the
FACETS simulator review (Brette et al. 2007) [2]_. These benchmarks are based
on the Vogels & Abbott network model [1]_ and are designed to test different
aspects of neural network simulation:

- **Benchmark 1 (COBA)**: Conductance-based synapses with integrate-and-fire neurons
- **Benchmark 2 (CUBA)**: Current-based synapses with integrate-and-fire neurons
- **Benchmark 3 (HH-COBA)**: Conductance-based synapses with Hodgkin-Huxley neurons
- **Benchmark 4 (CUBA-PS)**: Current-based synapses with precise spiking
- **Benchmark 5 (CUBA-STDP)**: Current-based synapses with STDP plasticity

All benchmarks create sparsely coupled networks of excitatory and inhibitory
neurons which exhibit self-sustained activity after an initial stimulus.
Connections within and across both populations are created at random. Both
neuron populations receive Poissonian background input.

Files
-----

- ``brette_et_al_2007_benchmark.py``: Shared framework module containing common
  network building and simulation functions
- ``coba.py``: Benchmark 1 - Conductance-based synapses (COBA)
- ``cuba.py``: Benchmark 2 - Current-based synapses (CUBA)
- ``hh_coba.py``: Benchmark 3 - Hodgkin-Huxley neurons (HH-COBA)
- ``cuba_ps.py``: Benchmark 4 - Precise spiking (CUBA-PS)
- ``cuba_stdp.py``: Benchmark 5 - STDP plasticity (CUBA-STDP)

Usage
-----

Each benchmark can be run directly as a Python script. For example::

    python coba.py

Or from the examples directory::

    python -m brette_et_al_2007.coba

The benchmarks will print a summary including:
- Number of neurons and synapses
- Average firing rates for excitatory and inhibitory populations
- Building and simulation times

Benchmark Details
-----------------

**Benchmark 1 (COBA)**
  - Neuron model: ``iaf_cond_exp``
  - Synapse model: conductance-based (exponential)
  - Spike times: grid-constrained
  - Simulation time: 1000 ms

**Benchmark 2 (CUBA)**
  - Neuron model: ``iaf_psc_exp``
  - Synapse model: current-based (exponential)
  - Spike times: grid-constrained
  - Simulation time: 10000 ms (10x longer due to lower computational load)

**Benchmark 3 (HH-COBA)**
  - Neuron model: ``hh_cond_exp_traub``
  - Synapse model: conductance-based (exponential)
  - Spike times: grid-constrained
  - Simulation time: 1000 ms

**Benchmark 4 (CUBA-PS)**
  - Neuron model: ``iaf_psc_delta``
  - Synapse model: current-based (delta/voltage jump)
  - Spike times: off-grid (precise spiking)
  - Simulation time: 10000 ms

**Benchmark 5 (CUBA-STDP)**
  - Neuron model: ``iaf_psc_exp``
  - Synapse model: STDP (E->E), static current (others)
  - Spike times: grid-constrained
  - Simulation time: 2000 ms
  - Features: Random initial weights and membrane potentials, STDP plasticity

References
----------

.. [1] Vogels TP, Abbott LF. 2005. Signal propagation and logic gating in
       networks of integrate-and-fire neurons. Journal of Neuroscience.
       25(46):10786-10795.
       https://doi.org/10.1523/JNEUROSCI.3508-05.2005

.. [2] Brette R, Rudolph M, Carnevale T, Hines M, Beeman D, Bower JM, et al.
       2007. Simulation of networks of spiking neurons: a review of tools and
       strategies. Journal of Computational Neuroscience. 23(3):349-398.
       https://doi.org/10.1007/s10827-007-0038-6



.. raw:: html

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

.. thumbnail-parent-div-open

.. raw:: html

    <div class="sphx-glr-thumbcontainer" tooltip="This script creates a sparsely coupled network of excitatory and inhibitory neurons which exhibits self-sustained activity after an initial stimulus. Connections within and across both populations are created at random. Both neuron populations receive Poissonian background input. The spike output of 500 neurons from each population are recorded. Neurons are modeled as leaky integrate-and-fire neurons with conductance-based synapses (exponential functions). The model is based on the Vogels &amp; Abbott network model [1]_.">

.. only:: html

  .. image:: /auto_examples/brette_et_al_2007/images/thumb/sphx_glr_coba_thumb.png
    :alt:

  :doc:`/auto_examples/brette_et_al_2007/coba`

.. raw:: html

      <div class="sphx-glr-thumbnail-title">Benchmark 1 of the simulator review (COBA)</div>
    </div>


.. raw:: html

    <div class="sphx-glr-thumbcontainer" tooltip="This script creates a sparsely coupled network of excitatory and inhibitory neurons which exhibits self-sustained activity after an initial stimulus. Connections within and across both populations are created at random. Both neuron populations receive Poissonian background input. The spike output of 500 neurons from each population are recorded. Neurons are modeled as leaky integrate-and-fire neurons with voltage jump synapses. Spike times are not constrained to the discrete time grid. The model is based on the Vogels &amp; Abbott network model [1]_.">

.. only:: html

  .. image:: /auto_examples/brette_et_al_2007/images/thumb/sphx_glr_cuba_ps_thumb.png
    :alt:

  :doc:`/auto_examples/brette_et_al_2007/cuba_ps`

.. raw:: html

      <div class="sphx-glr-thumbnail-title">Benchmark 4 of the simulator review (CUBA-PS)</div>
    </div>


.. raw:: html

    <div class="sphx-glr-thumbcontainer" tooltip="This script creates a sparsely coupled network of excitatory and inhibitory neurons which exhibits self-sustained activity after an initial stimulus. Connections within and across both populations are created at random. Both neuron populations receive Poissonian background input. The spike output of 500 neurons from each population are recorded. Neurons are modeled as leaky integrate-and-fire neurons with current-based synapses (exponential functions). The model is based on the Vogels &amp; Abbott network model [1]_.">

.. only:: html

  .. image:: /auto_examples/brette_et_al_2007/images/thumb/sphx_glr_cuba_thumb.png
    :alt:

  :doc:`/auto_examples/brette_et_al_2007/cuba`

.. raw:: html

      <div class="sphx-glr-thumbnail-title">Benchmark 2 of the simulator review (CUBA)</div>
    </div>


.. raw:: html

    <div class="sphx-glr-thumbcontainer" tooltip="This script creates a sparsely coupled network of excitatory and inhibitory neurons which exhibits self-sustained activity after an initial stimulus. Connections within and across both populations are created at random. Both neuron populations receive Poissonian background input. The spike output of 500 neurons from each population are recorded. Neurons are modeled as Hodgkin-Huxley neurons with conductance-based synapses (exponential functions). The model is based on the Vogels &amp; Abbott network model [1]_.">

.. only:: html

  .. image:: /auto_examples/brette_et_al_2007/images/thumb/sphx_glr_hh_coba_thumb.png
    :alt:

  :doc:`/auto_examples/brette_et_al_2007/hh_coba`

.. raw:: html

      <div class="sphx-glr-thumbnail-title">Benchmark 3 of the simulator review (HH-COBA)</div>
    </div>


.. raw:: html

    <div class="sphx-glr-thumbcontainer" tooltip="The fifth simulator review benchmark is implemented as a variation of the Brunel Network. This script creates a sparsely coupled network of excitatory and inhibitory neurons. Connections within and across both populations are created at random. Both neuron populations receive Poisson background input. The spike output of 500 neurons are recorded. Neurons are modeled as leaky integrate-and-fire neurons with current-injecting synapses (exponential functions). Excitatory-excitatory synapses implement multiplicative STDP.">

.. only:: html

  .. image:: /auto_examples/brette_et_al_2007/images/thumb/sphx_glr_cuba_stdp_thumb.png
    :alt:

  :doc:`/auto_examples/brette_et_al_2007/cuba_stdp`

.. raw:: html

      <div class="sphx-glr-thumbnail-title">Benchmark 5 of the simulator review (CUBA-STDP)</div>
    </div>


.. raw:: html

    <div class="sphx-glr-thumbcontainer" tooltip="This module provides a common framework for running the Brette et al. 2007 simulator review benchmarks. The benchmarks create sparsely coupled networks of excitatory and inhibitory neurons which exhibit self-sustained activity after an initial stimulus. The model is based on the Vogels &amp; Abbott network model [1]_.">

.. only:: html

  .. image:: /auto_examples/brette_et_al_2007/images/thumb/sphx_glr_brette_et_al_2007_benchmark_thumb.png
    :alt:

  :doc:`/auto_examples/brette_et_al_2007/brette_et_al_2007_benchmark`

.. raw:: html

      <div class="sphx-glr-thumbnail-title">Brette et al. 2007 Benchmark Framework</div>
    </div>


.. thumbnail-parent-div-close

.. raw:: html

    </div>


.. toctree::
   :hidden:

   /auto_examples/brette_et_al_2007/coba
   /auto_examples/brette_et_al_2007/cuba_ps
   /auto_examples/brette_et_al_2007/cuba
   /auto_examples/brette_et_al_2007/hh_coba
   /auto_examples/brette_et_al_2007/cuba_stdp
   /auto_examples/brette_et_al_2007/brette_et_al_2007_benchmark