mse.BW {landpred}

Helper function for optimize.mse.BW.

Description

Helper function for optimize.mse.BW.

Usage

mse.BW(data, t0,tau,h, folds = 3,reps=2)

Arguments

data

n by 5 matrix. A data matrix where the first column is XL = min(TL, C) where TL is the time of the long term event, C is the censoring time, and the second column is DL =1*(TL<C), the third column is XS = min(TS, C) where TS is the time of the short term event, C is the censoring time, the fourth column is DS =1*(TS<C), and the fifth column is the covariate. These are the data used to calculate the estimated probability.

t0

the landmark time.

tau

the residual survival time of interest.

h

bandwidth

folds

Number of folds wanted for K-fold cross-validation. Default is 3.

reps

Number of repitions wanted of K-fold cross-validation. Default is 2.

Value

mean of MSE

Author(s)

Layla Parast

References

Parast, Layla, Su-Chun Cheng, and Tianxi Cai. Incorporating short-term outcome information to predict long-term survival with discrete markers. Biometrical Journal 53.2 (2011): 294-307.