TITLE(avas @@ avas: Additivity and variance stabilization for regression )
USAGE(
avas(x, y, wt, mon, lin, cat, circ, delrsq,yspan)

)
ARGUMENTS(
ARG(x@@)
a matrix containing the independent variables.
ARG(y@@)
a vector containing the response variable.
ARG(wt@@)
an optional vector of weights.
ARG(mon@@)
an optional integer vector specifying which variables are to be
transformed by monotone transformations.
Positive values in LANG(mon) refer to columns of the LANG(x) matrix and zero
to the response variable.
ARG(lin@@)
an optional integer vector specifying which variables are to be
transformed by linear transformations.
Positive values in LANG(lin) refer to columns of the LANG(x) matrix and zero
to the response variable.
ARG(cat@@)
an optional integer vector specifying which variables assume
categorical values.
Positive values in LANG(cat) refer to columns of the LANG(x)
matrix and zero to the response variable.
ARG(circ@@)
an integer vector specifying which variables assume
circular (periodic) values.
Positive values in LANG(circ) refer to columns of the LANG(x)
matrix and zero to the response variable.
ARG(delrsq@@)
termination threshold.  Iteration stops when R-squared
changes by less than LANG(delrsq) in 3 consecutive iterations
(default 0.01).
ARG(yspan@@)
Optional window size parameter for smoothing the variance. 
Range[0,1]. Default=0 (cross validated choice). .5 is a reasonable
alternative to try.
)
PARA BOLD(VALUE)
structure with the following components:
ARG(x@@)
the input x matrix.
ARG(y@@)
the input y vector.
ARG(tx@@)
the transformed x values.
ARG(ty@@)
the transformed y values.
ARG(rsq@@)
the multiple R-squared value for the transformed values.
ARG(l@@)
not used in this version of avas
ARG(m@@)
not used in this version of avas
ARG(yspan@@)
span used for smoothing the variance
ARG(iters@@)
iteration number and rsq for that iteration
ARG(niters@@)
number of iterations used
PARA BOLD(REFERENCE)
Rob Tibshirani, "Estimating optimal transformations for regression". 
JASA Vol 83, No 402 Page 394
Version: Mar 18/1987
BLANK

PARA BOLD( )
BLANK

EXAMPLES(
TWOPI <- 8*atan(1)
x <- runif(200,0,TWOPI)
y <- exp(sin(x)+rnorm(200)/2)
a <- avas(x,y)
plot(a\$y,a\$ty)  # view the response transformation
plot(a\$x,a\$tx)  # view the carrier transformation
plot(a\$tx,a\$ty) # examine the linearity of the fitted model
)
PARA ITALIC(File automatically converted from S(-PLUS) help format)
