MATLAB: Why I can’t run LinearModel.fit on the Matlab R2011b

linearmodel.fit

I read the example of time series regression on http://www.mathworks.se/help/econ/examples/time-series-regression-i-linear-models.html . But I couldn't run LinearModel.fit on my matlab. I searched it in computer, and I didn't find it. It seems that this m-file should be included in the matlab, but what is the problem?

Best Answer

LinearModel.fit was introduced in R2012a of Statistics Toolbox.

Alan Weiss

MATLAB mathematical toolbox documentation

Related Solutions

MATLAB: How to calculate AIC in glmfit

It is recommended to use fitglm instead of the older glmfit. Depending on the release of MATLAB (> R2012a) you can use either of the two:

http://www.mathworks.com/help/stats/fitglm.html

http://www.mathworks.com/help/stats/generalizedlinearmodel.fit.html

You can access AIC, BIC etc as follows:

load hospital
modelspec = 'Smoker ~ Age*Weight*Sex - Age:Weight:Sex';
mdl = fitglm(hospital,modelspec,'Distribution','binomial')
mdl.ModelCriterion
ans = 
     AIC: 137.141380948166
    AICc: 138.358772252513
     BIC: 155.377572250082
    CAIC: 162.377572250082

MATLAB: How can i use aic to regression model and time series models matlab R2006a

AICBIC function is available in the Econometrics Toolbox:

http://www.mathworks.com/help/econ/aicbic.html

If you don't have the toolbox you won't be able to call the function. However if you know what the likelihood then AIC is is:

AIC = 2k - 2*ln(L)

http://en.wikipedia.org/wiki/Akaike_information_criterion

To check what toolboxes you have installed run:

>> ver

Best Answer

Related Solutions

MATLAB: How to calculate AIC in glmfit

MATLAB: How can i use aic to regression model and time series models matlab R2006a

Related Question