www.neuron.yale.edu

On 3/31/2005 a NEURON user aked:
Can a network model contain two different kinds of cells?

It can contain as many different kinds as you like.
If your net has only a few cells, you can do the whole thing
with the Network Builder (see tutorial at
http://www.neuron.yale.edu/neuron/docs
). If your net has > 5 or 6 cells and many synaptic connections,
use the approach described in chapter 11 of the NEURON book
http://www.neuron.yale.edu/ftp/ted/book ... xedref.pdf

If you are trying to make a network with multiple types of cells and you want the number in each cell class to vary, how would you do that? What would be the easiest way to write that kind of code?

Use for loops with different parameters.

The NetWork Builder tosses all biophysical model cells into a single big bag (a List of
objects, each of which is a specific instance of a particular cell class), and has another
big bag for all artificial spiking cells. I find it convenient to use a separate List for each
different cell class.

NUM_PYRAMIDS = 50
NUM_SPSTELLATE = 100

objref pyrlist, splist
for ii = 0, NUM_PYRAMIDS-1 pyrlist.append(new Pyramid())
for ii = 0, NUM_SPSTELLATE-1 splist.append(new SpinyStellate())