cut_dat {PWEXP}

Cut Data before a Specified Time

Description

Take a subset of a dataset by constraining the randomization time <= cut time. Additionally, it updates the follow-up time, censor/event indicator, censor reason, accordingly.

Usage

cut_dat(cut, data, var_randT=NULL, var_followT=NULL, var_followT_abs=NULL,
      var_censor=NULL, var_event=NULL, var_censor_reason='status_at_end')

Arguments

cut

cut time (from the beginning of the trial); only rows with randomization time <= cut will be kept.

data

a data frame.

var_randT

character; the variable name of randomization time. If missing, then the randomization time will be treated as 0 and NO subjects will be filtered by cut time.

var_followT

character; the variable name of follow-up time (from randomization)

var_followT_abs

character; the variable name of follow-up time (from the beginning of the trial)

var_censor

character; the variable name of censoring (drop-out or death) indicator (1=censor, 0=event)

var_event

character; the variable name of event indicator (1=event, 0=censor)

var_censor_reason

character; the variable name of censoring reason (character variable). This variable will be created, if data does not contain it.

Details

We first filter rows that randomization time is equal to or less then cut time. Then we modify these columns (if provided):

• var_followT: change values to (cut - randomization time) if (follow-up time + randomization time) > cut

• var_followT_abs: change values to cut if (follow-up time from beginning) > cut

• var_censor: change values to 1 if (follow-up time from beginning) > cut

• var_event: change values to 0 if (follow-up time from beginning) > cut

• var_censor_reason: change values to 'cut' if (follow-up time from beginning) > cut

Value

A subset data frame with the same columns as data.

var_censor_reason is the only variable that is allowed to be absent in data. The function will create this variable in the returned data frame and set values 'cut' to the subjects whose (follow-up time from beginning) > cut.

Note

The original dataset data will NOT be modified.

Author(s)

Tianchen Xu [e-mail]

Examples

event_dist <- function(n)rpwexp(n, rate = c(0.1, 0.01, 0.2), breakpoint =  c(5,14))
dat <- simdata(rand_rate = 20, total_sample = 1000, drop_rate = 0.03,
               advanced_dist = list(event_dist=event_dist),
               add_column = c('censor_reason','event','followT','followT_abs'))
cut <- quantile(dat$randT, 0.8)
train <- cut_dat(var_randT = 'randT', cut = cut, data = dat,
                 var_followT = 'followT', var_followT_abs = 'followT_abs',
                 var_event = 'event', var_censor_reason = 'censor_reason')