Plot and save survival performance of a model on test data — plot_survival

This function groups individuals into risk strata (e.g., Low/Medium/High) based on their predicted risk scores from a fitted survival model. It then plots Kaplan–Meier survival curves for each risk group, including a log-rank test for group separation and displays the concordance index (C-index) of the model on the test data.

Usage

plot_survival_performance(
  df_test,
  c_index = NULL,
  n_groups = 3,
  file_name = NULL
)

Arguments

df_test

A data frame containing at least the following columns:

time: Observed survival or follow-up time (numeric).
event: Event indicator (1 = event occurred, 0 = censored).
.pred: Predicted risk score or linear predictor from the model (higher values indicate higher risk).

c_index

Numeric. The concordance index (C-index) computed on the test data.

n_groups

Integer. Number of risk groups to stratify by (default = 3). Typically 3 groups correspond to "Low", "Medium", and "High" risk strata.

file_name

Character (optional). If provided, the Kaplan–Meier plot will be saved to "Results/Survival_KM_<file_name>.pdf".

Value

Invisibly returns the ggsurvplot object for further customization, and saves a PDF of the plot in the "Results/" directory if file_name is provided.

Details

Risk groups are defined by quantile-based cut points on the predicted risk scores. The log-rank test is used to assess whether survival curves differ significantly between risk strata. The C-index is displayed for interpretability.

Examples

if (FALSE) { # \dontrun{
plot_survival_performance(
  df_test = test_data,
  c_index = 0.74,
  n_groups = 3,
  file_name = "cox_model_test"
)
} # }