Paper_code

code for paper

Requried R packages

data.table >= 1.12.8
intervals >= 0.15.2

Code explanation

Explanation the code for ker_ts_poisson_mix.R

ker_ts_poisson_mix.R produces Figure 1 in the manuscript, Figure 3 in the Web Appendices, and Figure 1 and 3 in Respose to the comments.

Illustrate the main functions contained in ker_ts_poisson_mix.R code, as follows:

The defined "data.gen.mix()" function is to generate the simulation data set where the number of obaservation times is random, and recurrent event is generated from a mixed Poisson process;

The defined "data.gen()" function is to generate the simulation data set where the number of obaservation times is random, and recurrent event is generated from a Poisson process;

The defined "beta.fun()" and "beta.fun.ts()" functions are the Newton-Raphson iteration algorithm for the proposed method and the two-stage approach, respectively;

The defined "shape.fun()" function is the estimation of the shape function which is useful to estiamte the baseline function in the two-stage approach;

The defined "cvscore()" and "cvscore.ts()" functions are the cross-validation score function of the proposed method and the two-stage approach, respectively;

The defined "choose.hb()" and "choose.hb.ts()" functions are the cross-validation approach for bandwidth selection of the proposed method and the two-stage approach, respectively;

The defined "simulation.est()" and "simulation.ts()" are the simulation study for 500 replications of the proposed method and the two-satge approach, respectively;

The ploted figure saved as "cprandom.eps" produces Figure 3 in the Web Appendices, and Figure 3 in Respose to the comments;

The ploted figure saved as "cpmixpoisson.eps" produces Figure 1 in the Web Appendices, and Figure 1 in Respose to the comments.

Explanation the code for ker_ts_fixed.R

ker_ts_fixed.R produces Figure 5 in the Web Appendices, and Figure 4 and 6 in Respose to the comments.

Illustrate the main functions contained in ker_ts_fixed.R code, as follows:

The defined "data.gen()" function is to generate the simulation data set where the obaservation time is fixed;

The defined "beta.fun.ker()" and "beta.fun.ts()" functions are the Newton-Raphson iteration algorithm for the proposed method and the two-stage approach, respectively;

The defined "shape.fun()" function is the estimation of the shape function which is useful to estiamte the baseline function in the two-stage approach;

The defined "cvscore.ker()" and "cvscore.ts()" functions are the cross-validation score function of the proposed method and the two-stage approach, respectively;

The defined "choose.hb.ker()" and "choose.hb.ts()" functions are the cross-validation approach for bandwidth selection of the proposed method and the two-stage approach, respectively;

The defined "simulation.ker()" and "simulation.ts()" are the simulation study for 500 replications of the proposed method and the two-satge approach, respectively;

The ploted figure saved as "cpfixed.eps" produces Figure 5 in the Web Appendices, and Figure 4 in Respose to the comments;

The ploted figure saved as "fixkxq.eps" produces Figure 6 in Respose to the comments.

Explanation the code for ker_ts_heavytail.R

ker_ts_heavytail.R produces Figure 7 in the Web Appendices, and Figure 5 in Respose to the comments.

Illustrate the main functions contained in ker_ts_heavytail.R code, as follows:

The defined "data.gen()" function is to generate the simulation data set where the obaservation time is heavy-tailed;

The defined "beta.fun.ker()" and "beta.fun.ts()" functions are the Newton-Raphson iteration algorithm for the proposed method and the two-stage approach, respectively;

The defined "shape.fun()" function is the estimation of the shape function which is useful to estiamte the baseline function in the two-stage approach;

The defined "cvscore.ker()" and "cvscore.ts()" functions are the cross-validation score function of the proposed method and the two-stage approach, respectively;

The defined "choose.hb.ker()" and "choose.hb.ts()" functions are the cross-validation approach for bandwidth selection of the proposed method and the two-stage approach, respectively;

The defined "simulation.ker()" and "simulation.ts()" are the simulation study for 500 replications of the proposed method and the two-satge approach, respectively;

The ploted figure saved as "cpfixed.eps" produces Figure 7 in the Web Appendices, and Figure 5 in Respose to the comments.