Re: Multiple Correspondence Analysis

Previous Topic Next Topic
classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view

Re: Multiple Correspondence Analysis

Kevin Thorpe
You should send this to [hidden email].

On 03/09/2012 09:21 AM, Andrea Sica wrote:

> Hello everybody, I'm looking for someone who is able with MCA and
> would like to gives some help.
> If what I'm doing is not wrong, according to the purpose I have, I
> need to understand how to create a dependence matrix, where I can
> analyze the
> dependence between all my variables.
> Till now this is what I was able to do:
> /p <- length(spain)/ #this is the number of the variables (91)
> /chisquare <- matrix(spain, nrow=(p-1), ncol=p)/ #it creates a
> squared-matrix with all the variables (if I'm not already wrong)
> /for(i in (1:(p-1))){/
> /chisquare[i, (1:(p-1))] <- chisq.test(spain[,i], spain[, i+1])$statistic/
> /chisquare[i, p] <- chisq.test(spain[,i], spain[, i+1])$p.value/
> /} /#it should have related the "p" variables to analyze whether in
> pairs they are dependents, but it seems like it just related two of
> them and repeated the relations for all the number of columns (since
> it gives the same values in each cell by row)
> /chisquare/ #all the cells have the same values by row
> Anyway, I think is also the way I'm proceeding which is wrong, since I
> want to relate all the variables in pairs thus to be able to calculate
> the dependence between all of them. That's why I am going for a
> dependence matrix. Where am I wrong?
> After that I can proceed with the MCA. Of course, I would also
> need help there.
> I used the following codes to do it:
> /spain.mca <- mjca(spain) /#it makes the mca for all the data
> /spain.mca/
> /plot(spain.mca)/ #it shows the plot
> But the plot was overcrowded. Anyway, I must first complete the first
> step, this was just to make some practice on it.
> As you can see, until now I didn't succeed.
> I hope someone will be so gentle to give it a try. Attached you are
> the data-set
> Thank you
> Best

Kevin E. Thorpe
Biostatistician/Trialist,  Applied Health Research Centre (AHRC)
Li Ka Shing Knowledge Institute of St. Michael's
Assistant Professor, Dalla Lana School of Public Health
University of Toronto
email: [hidden email]  Tel: 416.864.5776  Fax: 416.864.3016

        [[alternative HTML version deleted]]

[hidden email] mailing list
PLEASE do read the posting guide
and provide commented, minimal, self-contained, reproducible code.