ftlim multiple regression with interaction term

Hi, I am trying to use ftlim to carry out multiple regression with an interaction term. I have looked at the example in mathworks and copied the example into my comand window. However, I keep getting an error "Predictor and response variables must have the same length". I have had the same problem with my own data. I am usin gMatlab R2013b. This is the example: http://www.mathworks.co.uk/help/stats/group-comparisons-using-categorical-arrays.html?refresh=true#zmw57dd0e3149 Can anyone help?

Thanks in advance.

[the cyclist]

Can you please confirm that if you type just these lines into the command window, that you get the error?

clear
load('carsmall')
cars = table(MPG,Weight,Model_Year);
cars.Model_Year = nominal(cars.Model_Year);
fit = fitlm(cars,'MPG~Weight*Model_Year')

and then what about this?

clear
rng(1)
X = randn(5,1);
Y = randn(5,1);
xy_table = table(X,Y);
fit = fitlm(xy_table,'Y~X')

[Aoife]

Hi, thanks for your response. Yes, I get the same error using moth of those command lines. This is the error: Error using classreg.regr.FitObject/assignData (line 257) Predictor and response variables must have the same length.

Error in classreg.regr.TermsRegression/assignData (line 349) model = [e-mail](model,X,y,w,asCat,varNames,excl);

Error in LinearModel.fit (line 852) model = assignData(model,X,y,weights,asCatVar,dummyCoding,model.Formula.VariableNames,exclude);

Error in fitlm (line 111) model = Linear

[Aoife]

I have just used the latest version of Matlab and no longer receive this error.Thanks for your help.

[the cyclist]

Glad it worked out.

Answer 1

[Sven]

It's a few years late but I think I've discovered the bug that may have been your problem (or at least generates a similar error that others might find via a search):

Linear regression (specifically the fitlm() method) fails when using a table as first input AND when that table contains a variable with 3 or more dimensions. This failure occurs even when that variable is not included as a term in the fit.

Bug is reproducible via the following, showing a fit working with minimal variables, still working with an unused 2d variable, and then failing with an unused 3d variable:

% Make a table 
X = load('carsmall') 
cars = table(X.MPG,X.Weight,nominal(X.Model_Year),'Var',{'MPG','Weight','Model_Year'}) 
% Show that fitlm works 
fitRunsOk = fitlm(cars,'MPG~Weight*Model_Year') 
cars.unused2dVar = rand(size(cars,1),5) 
fitStillOk = fitlm(cars,'MPG~Weight*Model_Year') 
% But it fails when an irrelevant variable is added that happens to be 3d 
cars.unused3dVar = rand(size(cars,1),5,5) 
fitFAILS = fitlm(cars,'MPG~Weight*Model_Year') 

>> Error using classreg.regr.FitObject/assignData (line 134) 
Predictor and response variables must have the same length.

The workaround until fixed (I've submitted a bug report) will be to make a temporary copy of your table that omits any variables with 3 or more dimensions.