[R] Multidimensional scaling on a correlation matrix
Holger.steinmetz at web.de
Fri Oct 30 16:08:54 CET 2009
when conducting a multidimensional scaling analysis, it is possible to do
this on a correlation matrix?
I calculated a dissimilarity matrix by substracting the corr.matrix from 1.
Then I calculated the distance matrix with d <- dist(dissimilarity matrix)
and ran the isoMDS function.
Is that correct?
In addition: The stress value was 2.9. I read that this is a percentage. So
- according to the usual manner to report stress values from 0 to 1, this
would be .029?
Finally, I compared the solution (diagram) with the alscal-algorithm from
SPSS (however computed on the raw data) and there were some slight
differences. Can you tell me the differences between alscal and isoMDS?
As you might see, I am not very experienced in MDS. Thus, any recommendation
(literatur etc.) would be a great help.
View this message in context: http://old.nabble.com/Multidimensional-scaling-on-a-correlation-matrix-tp26126667p26126667.html
Sent from the R help mailing list archive at Nabble.com.
More information about the R-help