LegacyArguments | R Documentation |
plot
and Effect
MethodsPrior to verson 4.0-0 of the effects package, there were many (literally dozens) of
arguments to the plot
methods for "eff"
and "effpoly"
objects.
In version 4.0-0 of the package, we have consolidated these arguments into a much smaller
number of arguments (e.g., lines
, points
, axes
) that take lists of
specifications. We have similarly consolidated some of the arguments to Effect
methods
into the confint
and fixed.predictors
arguments.
For backwards compatibility, we have to the extent possible retained the older arguments. If specified, these legacy arguments take precedence over the newer list-style arguments
Here is the correspondence between the old and new arguments.
For plot
methods:
multiline=TRUE/FALSE
lines=list(multiline=TRUE/FALSE)
type=c("rescale", "link", "response")
For models with a link function, "link"
plots in linear predictor scale, "response"
plots in the response scale, and the default "rescale"
plots in linear predictor scale but labels tick-marks in response scale.
z.var=which.min(levels)
lines=list(z.var=which.min(levels))
relevant only when lines=list(multiline=TRUE)
colors={vector of colors}
lines=list(col={vector of colors})
lty={vector of line types}
lines=list(lty={vector of line types})
lwd={vector of line widths}
lines=list(lwd={vector of line widths})
use.splines=TRUE/FALSE
lines=list(splines=TRUE/FALSE)
cex={number}
points=list(cex={number})
rug=TRUE/FALSE
axes=list(x=list(rug=TRUE/FALSE)
xlab={"axis title"}
axes=list(x=list(lab={"axis title"}))
xlim={c(min, max)}
axes=list(x=list(lim={c(min, max)}))
rotx={degrees}
axes=list(x=list(rot={degrees}))
ticks.x=list({tick specifications})
axes=list(x=list(ticks=list({tick specifications})))
transform.x=list(link={function}, inverse={function})
axes=list(x=list(transform=list({lists of transformations by predictors})))
ylab={"axis title"}
axes=list(y=list(lab={"axis title"}))
ylim={c(min, max)}
axes=list(y=list(lim={c(min, max)}))
roty={degrees}
axes=list(y=list(rot={degrees}))
ticks=list({tick specifications})
axes=list(y=list(ticks=list({tick specifications})))
alternating=TRUE/FALSE
axes=list(alternating=TRUE/FALSE)
grid=TRUE/FALSE
axes=list(grid=TRUE/FALSE)
ci.style="bands"/"lines"/"bars"/"none"
confint=list(style="bands"/"lines"/"bars"/"none"
)
band.transparency={number}
confint=list(alpha={number})
band.colors={vector of colors}
confint=list(col={vector of colors})
residuals.color={color}
partial.residuals=list(col={color})
residuals.pch={plotting character}
partial.residuals=list(pch={plotting character})
residuals.cex={number}
partial.residuals=list(cex={number})
smooth.residuals=TRUE/FALSE
partial.residuals=list(smooth=TRUE/FALSE)
residuals.smooth.color={color}
partial.residuals=list(smooth.col={color})
span={number}
partial.residuals=list(span={number})
show.fitted=TRUE/FALSE
partial.residuals=list(fitted=TRUE/FALSE)
factor.names=TRUE/FALSE
lattice=list(strip=list(factor.names=TRUE/FALSE))
show.strip.values=TRUE/FALSE
lattice=list(strip=list(values=TRUE/FALSE))
layout={lattice layout}
lattice=list(layout={lattice layout})
key.args={lattice key args}
lattice=list(key.args={lattice key args})
style="lines"/"stacked"
for plot.effpoly
, axes=list(y=list(style="lines"/"stacked"))
rescale.axis=TRUE/FALSE
type="rescale"/"response"/"link"
For Effect
methods:
confint=TRUE/FALSE
or a listmay be substituted for the se
argument.
confidence.level={number}
se=list(level={number})
given.values={named vector}
fixed.predictors=list(given.values={named vector})
typical={function}
fixed.predictors=list(typical={function})
offset={function}
fixed.predictors=list(offset={function})
partial.residuals=TRUE/FALSE
residuals=TRUE/FALSE
transformation
This argument to Effect
is not needed to compute effects. It can now be set directly with the plot
method with the argument axes = list(y = list(transformation=specification))
.
John Fox jfox@mcmaster.ca
Effect
, plot.eff
, plot.effpoly
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.