from
NESim
by Chris Eliasmith General package for large-scale biologically plausible simulations (with GUI).
NESim Help
NESim Help
Neural Engineering Simulator (NESim) Help
Examples
There are several examples that can be found in the NeuronData directory
of the main NESim directory. These are described briefly below.
1. SingleEnsemble
This example is the simplest possible system you can define in NESim.
It consists of a single ensemble with one input. It is described in detail
here.
2. SimplestCircuit
This example is a simple communication channel. An encoded scalar signal
is sent from one population to another.
3. SimpleFunction
This example demonstrates the kinds of functions that can be computed.
It is described in detail here.
4. SimpleFunction2
This example demonstrates how multiple functions can be decoded from a
single population (W).
5. SimpleHighD
This demonstrates a slightly more complex system that includes 4 populations.
One of those represents a 6-dimensional space and is driven by two other
3-dimensional populations.
6. AddVectors
This demonstrates simple 2D vector addition.
7. Multiplication
This shows how to implement the multiplication of two variables. This
demonstrates techniques useful for computing nonlinear transformations.
8. Polynomial
This shows how to compute polynomials of an encoded signal. It is described
in detail here.
9. Polynomial2
This shows how to compute two different polynomials from one population
(basically a combination of 8. and 4. above).
10. Integration
This implements a simple neural integrator. This is not a particularly
good integrator, but it allows you to play with the various simulation
parameters, and try different populations of neurons to see the effects
that these have. The external signal should be a step functions.
11. ControlledIntegration
This demonstrates a controllable neural integrator. The comments at the
beginning of the .txt file describe how to run this example.
12. Oscillator
This demonstrates a simple oscillator. The external signals should
be zero and a step input. Because this is a numerical oscillator, the
amplitude will increase exponentially. To minimize this effect, the
step size can be reduced. However, the neural solution (though not the
direct solution) will reach a maximum amplitude of 1 because it can't
represent values larger than this.
13. Otolith
This is a more complex example that is not included with the standard
NESim code. It can be downloaded here
(33MB). This simulation includes about 5000 neurons that perform a computation
that is thought to take place in parts of the vestibular system (see
the book
for details).