gpr_based: Confidence interval based on Gaussian process regression

In zhimeir/cfsurvival: A package to construct predictive confidence interval for survival data

Confidence interval based on Gaussian process regression

Description

Construct conformal predictive interval based on distributional boosting

Usage

gpr_based(x, c, alpha, data_fit, data_calib, weight_calib, weight_new)

Arguments

x	a vector of the covariate of the test data.
c	the censoring time of the test data.
alpha	a number betweeo 0 and 1, specifying the miscaverage rate.
data_fit	a data frame, containing the training data.
data_calib	a data frame, containing the calibration data.
type	either "marginal" or "local". Determines the type of confidence interval.
dist	The distribution of T used in the cox model.

Value

low_ci a value of the lower bound for the survival time of the test point.

includeR 0 or 1, indicating if [r,inf) is included in the confidence interval.

See Also

Other model: aft_based(), distBoost_based(), np_based(), ph_based(), portnoy_based(), pow_based(), quantBoost_based(), rf_based()

zhimeir/cfsurvival documentation built on April 13, 2022, 6:41 a.m.