Predict the centred dose-response curve at a reference covariate setting

Returns the linear predictor of a fitted fit_aft_spline() model on an exposure grid, holding the covariates fixed at v_ref. Subtracting the first element produces the centred curve used throughout the manuscript.

Usage

predict_curve(fit, x_grid, v_ref = NULL, x_var)

Arguments

fit

A survreg object returned by fit_aft_spline().

x_grid

Numeric vector of exposure values to predict at.

v_ref

Named numeric vector of reference covariate values, with names matching the covariates used when fitting. Pass NULL if the model has no covariates.

x_var

Name of the exposure column (must match fit_aft_spline).

Value

A numeric vector of length length(x_grid) (the uncentred linear predictor; subtract out[1] to centre at the lower grid boundary).

Examples

sim <- generate_aft_data(n = 300, n_val = 100, seed = 1)
fit <- fit_aft_spline(sim$survival, x_var = "X_true",
                      covariates = c("V1","V2","V3","V4"))
x_grid <- seq(7, 13, length.out = 50)
lp <- predict_curve(fit, x_grid,
                    v_ref = c(V1=30, V2=30, V3=0, V4=0),
                    x_var = "X_true")