Build and cross validate a piecewise exponential model.
cv.pwexpm_fit(time, event, nfold=5, nsim=100, breakpoint=NULL,
nbreak=0, exclude_int=NULL, min_pt_tail=5, max_set=1000, seed=1818,
optimizer='mle', tol=1e-4, parallel=FALSE, mc.core=4, ...)A object of class "cv.pwexpm" is a numeric vector containing the CV log likelihood in each round of simulation. The plot function can be used to make a boxplot of the CV log likelihoods from pwexpm.
observed time from randomization.
the status indicator. See pwexpm_fit.
the number of folds used in CV.
the number of simulations.
pre-specified breakpoints. See pwexpm_fit.
total number of breakpoints. See pwexpm_fit.
an interval that excludes any estimated breakpoints. See pwexpm_fit.
the minimum number of events used for estimating the tail (the hazard rate of the last piece). See pwexpm_fit.
maximum estimated combination of breakpoints. See pwexpm_fit.
a random seed. Do not set seed if seed=NULL.
one of the optimizers: mle, ols, or hybrid. See pwexpm_fit.
the minimum allowed gap between two breakpoints. The gap is calculated as (max(time)-min(time))*tol. Keep it as default in most cases.
logical. If TRUE, use doSNOW package to run in parallel.
number of processes allowed to be run in parallel.
internal function reserved.
Tianchen Xu zjph602xutianchen@gmail.com
Use cross validation obtain the prediction log likelihood.
cv.pwexpm
event_dist <- function(n)rpwexpm(n, rate = c(0.1, 0.01, 0.2), breakpoint = c(5,14))
dat <- simdata(rand_rate = 20, drop_rate = 0.03, total_sample = 1000,
advanced_dist = list(event_dist=event_dist),
add_column = c('censor_reason','event','followT','followT_abs'))
# here nsim=10 is for demo purpose, pls increase it in practice!!
# \donttest{
cv0 <- cv.pwexpm_fit(dat$followT, dat$event, nsim = 10, nbreak = 0)
cv1 <- cv.pwexpm_fit(dat$followT, dat$event, nsim = 10, nbreak = 1)
cv2 <- cv.pwexpm_fit(dat$followT, dat$event, nsim = 10, nbreak = 2)
sapply(list(cv0,cv1,cv2), median)
# }
Run the code above in your browser using DataLab