seglines.RdPlots the piecewise segmented curve made up of Wald statistics versus estimates, using a colour code for the HDE severity.
seglines(x, y, dy, ddy, lwd = 2, cex = 2, plot.it = TRUE,
add.legend = TRUE, cex.legend = 1,
position.legend = "topleft", eta0 = NA, COPS0 = NA,
lty.table = c("solid", "dashed", "solid", "dashed",
"solid", "dashed", "solid"),
col.table = rainbow.sky[-5], pch.table = 7:1,
severity.table = c("None", "Faint", "Weak", "Moderate",
"Strong", "Extreme", "Undetermined"), FYI = FALSE, ...)Same as hdeffsev.
Graphical parameters: line width, and character expansion.
Logical, plot it? If FALSE then the other
graphical arguments are ignored.
Logical and character; add a legend?
The other argument is fed
into legend.
Self-explanatory.
Same as hdeffsev.
Graphical parameters for the 7 different types of segments.
Usually users should not assign anything to these arguments.
Setting pch.table = NULL will
suppress pch symbols from the legend.
Should be ignored.
This function was written to
complement hdeffsev
and is rough-and-ready.
It plots the signed Wald statistics as a function of
the estimates, and uses a colour-code to indicate
the severity of the
Hauck-Donner effect (HDE).
This can be obtained from its first two derivatives.
This function returns the severity of the HDE, possibly invisibly.
This function is likely to change in the short future because it is experimental and far from complete.
deg <- 4 # myfun is a function that approximates the HDE
myfun <- function(x, deriv = 0) switch(as.character(deriv),
'0' = x^deg * exp(-x),
'1' = (deg * x^(deg-1) - x^deg) * exp(-x),
'2' = (deg * (deg-1) * x^(deg-2) - 2*deg * x^(deg-1) +
x^deg) * exp(-x))
if (FALSE) { # \dontrun{
curve(myfun, 0, 10, col = "white")
xgrid <- seq(0, 10, length = 101)
seglines(xgrid, myfun(xgrid), myfun(xgrid, deriv = 1),
COPS0 = 2,
myfun(xgrid, deriv = 2), pch.table = NULL,
position = "bottom")
} # }