Create A Probability Plot for a Parametric Survival Model
DESCRIPTION:
Create a probability plot for a parametric survival model with one or fewer
covariates or stratification variables.
USAGE:
probplot.censorReg(object, plot.at=unique(group.variable),
method=c("KM", "separate", "factor", "regression", "one", "none"),
xlab=paste("Time to Failure - ", object$threshold),
ylab=paste(distribution2, "Probability"), main.title=NULL, sub=NULL,
conf.level=0.95, xlim=NULL, ylim=NULL, mle=T, mle.interval=F, upper=T,
ower=T, pch.data=1:lobj, lty.mle=1:lobj, lty.interval=1:lobj,
col.data=1:lobj, col.mle=1:lobj, col.interval=1:lobj, add.legend=F,
data=NULL)
REQUIRED ARGUMENTS:
- object
-
an object of class
"censorReg"
or
"censorRegList"
containing one or
fewer covariates or stratification variables.
OPTIONAL ARGUMENTS:
- plot.at
-
the values of the covariate for which a probability distribution is to be
plotted. Ignored if there is no covariate in the model in object.
- method
-
character string specifying method for computing the quantiles of the plotted
points on the graph. In addition to probabilities for the model in
object
, probabilities for one other model can be plotted as points. Let
the covariate be
x
, and let the response be
censor(y)
. Then the possible
methods for these alternative models are:
"KM" censor(y) ~ strata(x) Turnbull's generalization of
Kaplan-Meier estimates
"separate" censor(y) ~ strata(x) censorReg estimates with separate
location and scale for each covariate
value.
"factor" censor(y) ~ factor(x) censorReg estimates with separate
location and the same scale for each
covariate value
"regression" censor(y) ~ x censorReg model with location predicted
by regression on the covariate
"one" censor(y) ~ 1 censorReg model with no covariates
"none" The comparison points are not plotted.
The default is method="KM".
- xlab
-
character string specifying the x-axis label.
- ylab
-
character string specifying the y-axis label.
- main.title
-
character string specifying the main title. If NULL, an appropriate main
title is generated. Use main.title="" for no main title.
- sub
-
character string specifying a sub title displayed below the x-axis label.
If not specified, and
date.it = T
this will default to the date the object
was created.
- conf.level
-
the confidence level for plotting confidence intervals of the estimated
model probabilities.
- xlim
-
numeric vector specifying the x-axis limits on the scale of the response.
- ylim
-
numeric vector specifying the y-axis limits on the probability scale.
- mle
-
logical value indicating whether or not to add the maximum likelihood
estimate(s) of the estimated distribution(s).
- mle.interval
-
logical value indicating whether or not to add confidence intervals around
the maximum likelihood estimate(s).
- upper
-
logical value indicating whether or not to add the upper confidence limit for
the maximum likelihood estimate(s).
- lower
-
logical value indicating whether or not to add the lower confidence limit for
the maximum likelihood estimate(s).
- pch.data
-
vector of indices for plot symbols. The i-th element in pch.data contains the
plot symbol index (see pch in function par) for the i-th probability
distribution. The plot symbol is used to display the alternative model
probabilities.
- lty.mle
-
vector of line types for the maximum likelihood estimate of the failure
probability. The i-th element in lty.mle contains the line type (see lty in
function par) for the i-th probability distribution.
- lty.interval
-
vector of line types for the maximum likelihood confidence interval estimates.
The i-th element in lty.interval contains the line type (see lty in
function par) for the i-th set of confidence intervals.
- col.data
-
vector of color indices for plot symbols. The i-th element in col.data contains
the color index (see col in function par) for the i-th probability distribution.
The color is used to display the alternative model probabilities.
- col.mle
-
vector of color types for the maximum likelihood estimate of the failure
probability. The i-th element in col.mle contains the color type (see col in
function par) for the i-th probability distribution.
- col.interval
-
vector of color types for the maximum likelihood confidence intervals. The i-th
element in col.interval contains the color type (see col in function par) for
the i-th set of confidence intervals.
- add.legend
-
if T, a legend is added to the plot (using the locator() function). The legend
gives the quantiles associated with each line and color type.
- legend.loc
-
location for placing the upper left corner of the legend. If missing, you must click in the graph to locate the legend. May also specify a vector of values for locating the legend or the string
"auto"
for automatic placement.
VALUE:
the values plotted for the alternative distribution(s).
DETAILS:
This function is a method for the generic function
plot
for class
"censorReg"
. It can be invoked by calling
probplot
for an object of
the appropriate class.
SIDE EFFECTS::
A probability plot of the fit in
object
is produced on the current graphics
device.
SEE ALSO:
censorReg, , qqplot.censorReg.setup, qqplot.censorReg.plot.quantiles
EXAMPLES:
a <- censorReg(censor(days,event) ~ voltage, data=capacitor2, weights = weights)
probplot.censorReg(a)
probplot.censorReg(a, plot.at=c(26,32), method="separate")