MATLAB: Multiple species using “classify”

classifylinear discriminant analysisstatistics

Hello,

I have been working with the classify command and have been playing around with the example for Fisher's Iris ("load fisheriris"). In the example, only two species are used (versicolor and virginica); however, there are three in the complete data set (includes setosa). Using these data, I have tried using "classify" for all three species at the same time and it treats it like two. Does "classify" support more than two data sets? If so, could you please suggest some modifying code?

Many thanks! Jennifer

Best Answer

What example do you mean? On this page:

http://www.mathworks.com/products/statistics/demos.html?file=/products/demos/shipping/stats/classdemo.html

the example seems to be using all three species.

Related Solutions

MATLAB: Finding non-constant species in Simbiology Object

What you get is not a function handle but an heterogenous array of SimBiology objects. This array contains a mix of different component types like species, parameters or potentially compartments.

If this is what you want, you can get their names with

names = get(VariableSpecies, 'Name')

But if you only wanted species (not parameters nor compartments), you can get them like this:

speciesObj = sbioselect(modelObj, 'Type', 'Species','Where','Constant','==',false)
names = get(speciesObj, 'Name')

MATLAB: Get initial amounts of some species

As of R2011a, you can use sbioparamestim to estimate the initial amount of a species. If you can't easily upgrade to R2011a, you can either try another estimation function (e.g., sbionlinfit) or work around the limitation by estimating a parameter that controls the initial amount via an initialAssignment rule.

-Arthur

P.S. Here's a simple example to estimate the initial amount of x in the radiodecay model:

% Create a parameter and initial assignment rule
sbioloadproject radiodecay.sbproj m1
x = sbioselect(m1, 'Name', 'x');
x0 = m1.addparameter('x0', 1000, 'ValueUnits', 'molecule');
m1.addrule('x = x0', 'initialAssignment');
% Create synthetic data
c = sbioselect(m1, 'Name', 'c');
t = [1:10]'; y = 2000*exp(-c.Value*t);
% Fit the data, simulate, and plot
[x0Estimate, result] = sbioparamestim(m1, t, y, x, x0)
x0.Value = x0Estimate;
sd = sbiosimulate(m1);
plot(sd.Time, sd.Data, '-', t, y, 'x')

Best Answer

Related Solutions

MATLAB: Finding non-constant species in Simbiology Object

MATLAB: Get initial amounts of some species

Related Question