survMisc (version 0.5.3)

predict: predicted events

Description

predicted events

Usage

"predict"(object, ..., eMP = TRUE, reCalc = FALSE)

Arguments

object

An object of class ten.

eMP

Add column(s) indicating events minus predicted.

...

Additional arguments (not implemented).

reCalc

Recalcuate the values? If reCalc=FALSE (the default) and the ten object already has the calculated values stored as an attribute, the value of the attribute is returned directly.

Value

An attribute, pred is added to object:

t

Times with at least one observation

P_

predicted number of events

And if eMP==TRUE (the default):

eMP_

events minus predicted

The names of the object's covariate groups are used to make the suffixes of the column names (i.e. after the _ character).

Details

With $K$ covariate groups, We use $ncg[i, k]$, the number at risk for group $k$, to calculate the number of expected events: $$P_{ik} = \frac{e_i(ncg_{ik})}{n_i} \quad k=1, 2 \ldots K$$

See Also

?survival::predict.coxph methods("predict")

Examples

## K&M. Example 7.2, Table 7.2, pp 209-210.
data("kidney", package="KMsurv")
k1 <- ten(Surv(time=time, event=delta) ~ type, data=kidney)
predict(k1)
predict(asWide(k1))
stopifnot(predict(asWide(k1))[, sum(eMP_1 + eMP_2)] <=
          .Machine$double.neg.eps)
## Three covariate groups
## K&M. Example 7.4, pp 212-214.
data("bmt", package="KMsurv")
b1 <- ten(Surv(time=t2, event=d3) ~ group, data=bmt)
predict(b1)
## one group only
predict(ten(Surv(time=t2, event=d3) ~ 1, data=bmt))