Skip to content

Manipulate a model

This document describes how to manipulate a xolotl model. Here, we assume you have a simple (one or a few compartments) model whose parameters you want to explore.

Manipulating a model out of the box

All xolotl models can be manipulated out of the box. If you run

x = xolotl.examples.neurons.BurstingNeuron;

a window will pop up with sliders for every parameter, and x.plot is used to generate plots of voltage and calcium dynamics that will auto-update as you move the sliders around.

To manipulate only a few parameters, you can call manipulate with specific arguments:

% only manipulate maximal conductances

% only manipulate a specific conductance

Writing a custom plot function

You can also write your own functions that will make a plot that will update as you vary the sliders created by x.manipulate. Your custom function should

  • integrate the model and interpret outputs
  • update or make plots as needed
  • not change any parameters of the xolotl object
  • should not have any outputs
  • should have only one input, a xolotl object

Let's assume your custom function is called foo(x). To configure xolotl to use your custom function, use this syntax:

x.manipulate_plot_func = {@x.plot};

Your custom function should look something like this:

% Your function should have no outputs, and only one input,
% which is a xolotl object

function foo(x)

% check if you need to make a figure. you can
% store handles to figures, and other data in
% x.handles
% this code snippet does the trick: 

if ~isfield(x.handles,'foo') ...
    || ~isvalid(')

    % need to make the figure

    % insert code to make the figure here

    % save your figure handles in x.handles


% now you have to integrate the model
[V,Ca] = x.integrate;

% this code will be called every time the slider
% is moved. so make it snappy, and update plots
% efficiently

% the best way to do this is to update
% the XData and YData of your saved handles
% to plots

To see an example of this working, look at this example file that you can run using:



The bigger your model is, the more sluggish manipulating it can appear. Try to use the simplest possible model to manipulate.

See Also