Calculate Residual Life Values Using a flexsurvreg() Object

Calculate residual life values using a fitted flexsurvreg object. Allows specifying covariate profiles via newdata.

Usage

reslifefsr(obj, life, p = 0.5, type = "mean", newdata = data.frame())

Arguments

obj: A fitted flexsurvreg object from package flexsurv.
life: Scalar numeric value of elapsed time at which to calculate residual life.
p: Percentile for quantile residual life. Default is 0.5.
type: Type of residual life to return: "mean", "median", "quantile", "percentile", or "all". Default is "mean".
newdata: A data frame containing covariate values to evaluate. Default is empty (evaluates baseline).

Value

A numeric vector (if type is "mean", "median", or "quantile") or a data frame (if type == "all"), with one row per covariate profile in newdata.

References

Jackson CH (2016). “flexsurv: a platform for parametric survival modeling in R.” Journal of statistical software, 70.

Poynor V (2010). “Bayesian inference for mean residual life functions in survival analysis.” Masters diss., University of California, Santa Cruz.

Prentice RL (1975). “Discrimination among some parametric models.” Biometrika, 62(3), 607–614.

Stacy EW (1962). “A generalization of the gamma distribution.” The Annals of mathematical statistics, pp. 1187–1192.

Examples

library(flexsurv)
fitg <- flexsurvreg(Surv(futime, fustat) ~ 1, data = ovarian, dist = "gengamma")
reslifefsr(obj = fitg, life = 6, p = 0.75, type = "all")
#>       mean   median percentile
#> 1 913.9498 903.1636   2856.237

fitg2 <- flexsurvreg(Surv(futime, fustat) ~ age, data = ovarian, dist = "gengamma")
df_new <- data.frame(age = c(50, 65))
reslifefsr(obj = fitg2, life = 3, type = "median", newdata = df_new)
#> [1] 1368.6166  366.1353