linear-pred-survival-metrics: Linear predictor survival metrics

In yardstick: Tidy Characterizations of Model Performance

Linear predictor survival metrics

Description

Linear predictor survival metrics evaluate survival model predictions based on linear predictors (log-hazard or log-risk scores).

Input requirements

• truth: a survival::Surv() object

• estimate: numeric (linear predictor values)

Available metrics

royston_survival()

Direction: maximize. Range: [0, 1]

See Also

dynamic-survival-metrics for time-dependent survival metrics

integrated-survival-metrics for integrated survival metrics

static-survival-metrics for static survival metrics

vignette("metric-types") for an overview of all metric types

Examples

data("lung_surv")

head(lung_surv)

lung_surv |>
  royston_survival(truth = surv_obj, estimate = .pred_linear_pred)

yardstick documentation built on April 8, 2026, 1:06 a.m.