========================
PyNN 0.9.2 release notes
========================

November 22nd 2017

Welcome to PyNN 0.9.2!


Recording injected currents
---------------------------

It is now possible to record the injected current from :class:`CurrentSource` objects in PyNN, for example:

.. code-block:: python

    noise = sim.NoisyCurrentSource(mean=0.5, stdev=0.2, start=50.0, stop=450.0, dt=1.0)
    noise.record()

    sim.run(500.0)

    signal = noise.get_data()

The returned signal object is a Neo :class:`AnalogSignal`.

Python 2.6
----------

As of this version, PyNN no longer supports Python 2.6.


NEST 2.14.0 and NEURON 7.5
--------------------------

PyNN 0.9.1 now supports the latest versions of NEST and NEURON.
NEURON 7.4 is also still supported.
NEST 2.12.0 should still work in most circumstances,
but current recording (see above) requires a more recent version.


native_electrode_type
---------------------

It has been possible for some time to use native (NEST-specific) neuron and synapse models with :mod:`pyNN.nest`.
It is now also possible to use native current generator models, e.g.:

.. code-block:: python

    noise = sim.native_electrode_type('noise_generator')(mean=500.0, std=200.0, start=50.0,
                                                         stop=450.0, dt=1.0)

Bug fixes
---------

A `number of bugs`_ have been fixed.

.. _`number of bugs`: https://github.com/NeuralEnsemble/PyNN/issues?q=is%3Aissue+milestone%3A0.9.2+is%3Aclosed