Hello R helpers!
I'm using the avas function form package acepack (called from areg.boot package Hmisc) to estimate automatically transformations of predictors (in this case monotonous) and response.
Well, it seems to work quite well, but I have 3 basic questions:
- which set of basis functions is used in this procedure?
- how do I back transform my estimate (y hat ) to the originasl scale?
- is there a closed analytical description of the model used? If so, how do I get it?
Thanks a lot
Anne
----------------------------------------------------
Anne Piotet
Email: anne.piotet@m-td.com
---------------------------------------------------
M-TD Modelling and Technology Development
PSE-C
CH-1015 Lausanne
Switzerland
Tel: +41 21 693 83 98
Fax: +41 21 646 41 33
--------------------------------------------------
[[alternative HTML version deleted]]