using transform.poly for multifactorial polynomial regression

[Andrew Czarn <aczarn@csse.uwa.edu.au>]

to anyone who can assist

i have fitted the following model using poly


*********************************************************************************************
tempPooled2.lm<-lm(generationFirstAt1.0E.10MeanPooled~(crValuePooledON+poly(muValuePooledON,2)
+experimentNumberF),data=dataFrameForAnalysisPooled)
*********************************************************************************************

the following are the coefficients obtained

*********************************************************************************************

coef(tempPooled2.lm)
 (Intercept) crValuePooledON poly(muValuePooledON, 2)1
    82.35894       -13.56899                 -21.75255

 poly(muValuePooledON, 2)2 experimentNumberF1 experimentNumberF2
                  40.80758        -0.07390909        -0.01648052

 experimentNumberF3 experimentNumberF4
        -0.01240909          0.0132974

*********************************************************************************************

i wish to use poly.transform to recover the simple polynomial form ie
obtain the simple polynomial form coefficients.

however, i get the following error message when trying the following:

poly.transform(poly(muValuePooledON,2),coef(tempPooled2.lm)[c(3,4)]))
Problem in poly.transform(poly(muValuePooledON, 2), ..: wrong length for 
coefs
*********************************************************************************************

can anyone please assist

thanks

andrew

Dr AST Czarn
University of Western Australia