ACE and AVAS
On Sun, 12 May 2002, Niels Waller wrote:
Hello, I would like to perform a monotone transformation of y (my response variable) to achieve additivity. My model is as follows y~X+items I do not want to transform X or items (both of these variables are factors)
X
[1] 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 4 4 4 4 4 4 4 4 4 5 5 5 5 5 5 5 5 5 6 6 6 6 6 6 6 6 6 7 7 7 7 7 7 [61] 7 7 7 8 8 8 8 8 8 8 8 8 Levels: 1 2 3 4 5 6 7 8
items
[1] 6 6 6 6 6 6 6 6 1 1 1 1 1 1 1 1 8 8 8 8 8 8 8 8 7 7 7 7 7 7 7 7 4 4 4 4 4 4 4 4 9 9 9 9 9 9 9 9 2 2 2 2 2 2 2 2 3 3 3 3 [61] 3 3 3 3 5 5 5 5 5 5 5 5 Levels: 1 2 3 4 5 6 7 8 9
y
[1] 0.00 0.00 0.00 0.00 0.00 0.00 0.27 0.00 0.73 0.06 0.04 0.04 0.19 0.06 0.23 0.51 0.21 0.64 0.07 0.15 0.08 0.39 0.18 0.33 [25] 0.61 0.52 0.67 0.18 0.24 0.12 0.40 0.52 0.51 0.64 0.68 0.79 0.13 0.33 0.30 0.51 0.73 0.68 0.68 0.84 0.78 0.13 0.28 0.64 [49] 0.58 0.95 0.91 0.77 0.97 0.78 0.17 0.47 0.85 0.82 1.00 0.93 0.90 0.97 0.90 0.17 0.85 1.00 0.98 1.00 1.00 1.00 1.00 1.00 I have several questions: (1) if I combine X and items by cbind(X,items) I loose their factor status. Obviously I should be doing something else.
ace() and avas() don't know about factors, and you can't put factors into matrices. If you construct a set of indicator variables you could put them into a matrix.
(2) I do not appear to be interpreting the help file correctly. I thought that if I only wanted to transform y then I would say XX<-cbind(X,items) ace(XX,y,mon=0)
No. All the variables are transformed. You probably want to use lin= to specify linear transformations, equivalent to regression without transforming. You'd probably be better off with avas() than ace() -- it's designed to give additivity and variance stabilisation, which seems to be what you want for that sort of regression model
(3) The function returns a value called $z - but this value is not documented. What is it?
Random junk from the fortran computations. That's why it's not documented. Yes, it probably should be removed. -thomas -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._