Hello everybody.
I have some Problems with the functions factanal() and
princomp(). Concerning factanal: What exactly is the difference between the
arguments type="regression" and type="weighted.ls"? In the first case, I
can't solve my Problem because of matrix singularity. In the second case
(type="weighted.ls"), I get a solution but I don't know how to interpret it.
Obviously, it is not the same as with type="regression".
It must be some kind of smoothing, but what kind?
What I need is the first princpal component of a set of variables, and the
data.frame of the
variables is obvously singular. I too tried out the function princomp(),
which works perfectly, but then I don't find the values of the principal
components. (In the Helpfile of princomp.objects, there is written, that I
would find the principal component in the first column of ..$loadings, but
this does'nt look like the principal component. Neither do the ...$scores. I
compared the loadings of the princomp object to the scores of the
factanal-objects and found out, that this things are not at all the same,
which they should aproximately be according to theory, doesen't they?
Thank you very much for your help. Ruth Meili
|