Dear list,
I have been trying to reproduce the scripts of the MASS library and I have
found a problem when trying to reproduce the code in 12.4, with the use of
neural networks and the function b1:
b1 <- function(Z, ...)
{
zp <- Z[,3] - pmax(Z[,2], Z[,1])
contour(xp/log(10), yp/log(10), matrix(zp, np),
add=T, levels=0, labex=0, ...)
zp <- Z[,1] - pmax(Z[,3], Z[,2])
contour(xp/log(10), yp/log(10), matrix(zp, np),
add=T, levels=0, labex=0, ...)
}
Using this function, the contours appear in a much reduced zone of the
chart. Instead, if I use
b1 <- function(Z, ...)
{
zp <- Z[,3] - pmax(Z[,2], Z[,1])
contour(exp(xp), exp(yp), matrix(zp, np),
add=T, levels=0, labex=0, ...)
zp <- Z[,1] - pmax(Z[,3], Z[,2])
contour(exp(xp), exp(yp), matrix(zp, np),
add=T, levels=0, labex=0, ...)
}
I get similar panels to the ones in the book. Also, this choice would be
consistent with the functions predplot and cushplot.
Am I doing something wrong?
Regards
Jordi
[[alternative HTML version deleted]]