predictCoxPL {riskRegression} | R Documentation |
Computation of survival probabilities from Cox regression models using the product limit estimator.
Description
Same as predictCox except that the survival is estimated using the product limit estimator.
Usage
predictCoxPL(
object,
times,
newdata = NULL,
type = c("cumhazard", "survival"),
keep.strata = TRUE,
keep.infoVar = FALSE,
...
)
Arguments
object |
The fitted Cox regression model object either
obtained with |
times |
[numeric vector] Time points at which to return the estimated hazard/cumulative hazard/survival. |
newdata |
[data.frame or data.table] Contain the values of the predictor variables
defining subject specific predictions.
Should have the same structure as the data set used to fit the |
type |
[character vector] the type of predicted value. Choices are
Several choices can be
combined in a vector of strings that match (no matter the case)
strings |
keep.strata |
[logical] If |
keep.infoVar |
[logical] For internal use. |
... |
additional arguments to be passed to |
Details
Note: the iid and standard errors are computed using the exponential approximation.
Examples
library(survival)
#### generate data ####
set.seed(10)
d <- sampleData(40,outcome="survival")
nd <- sampleData(4,outcome="survival")
d$time <- round(d$time,1)
#### Cox model ####
fit <- coxph(Surv(time,event)~ X1 + X2 + X6,
data=d, ties="breslow", x = TRUE, y = TRUE)
## exponential approximation
predictCox(fit, newdata = d, times = 1:5)
## product limit
predictCoxPL(fit, newdata = d, times = 1:5)
#### stratified Cox model ####
fitS <- coxph(Surv(time,event)~ X1 + strata(X2) + X6,
data=d, ties="breslow", x = TRUE, y = TRUE)
## exponential approximation
predictCox(fitS, newdata = d, times = 1:5)
## product limit
predictCoxPL(fitS, newdata = d, times = 1:5)
#### fully stratified Cox model ####
fitS <- coxph(Surv(time,event)~ 1,
data=d, ties="breslow", x = TRUE, y = TRUE)
## product limit
GS <- survfit(Surv(time,event)~1, data = d)
range(predictCoxPL(fitS)$survival - GS$surv)
fitS <- coxph(Surv(time,event)~ strata(X2),
data=d, ties="breslow", x = TRUE, y = TRUE)
## product limit
GS <- survfit(Surv(time,event)~X2, data = d)
range(predictCoxPL(fitS)$survival - GS$surv)