weibulltools (version 1.0.1)

mcs_mileage: Estimation of Driving Distances for Censored Observations using a Monte Carlo Approach

Description

This function simulates driving distances for censored observations under the condition that the operating time of these items is known up to a certain date where analysis is made. Operating times for these units can be estimated with functions like mcs_delay_register, mcs_delay_report and mcs_delays. The failed observations (where the driving distances are known) are used to estimate an annual mileage distribution. If the mileage distribution is fully specified annual random driving distances are drawn from this distribution and afterwards adjusted to the operating times of the censored observations.

Usage

mcs_mileage(x, event, mileage, distribution = "lognormal", seed = NULL,
  details = FALSE)

Arguments

x

a numeric vector of operating times. If not available use NA.

event

a vector of binary data (0 or 1) indicating whether unit i is a right censored observation (= 0) or a failure (= 1).

mileage

a numeric vector of driven distances. If not available use NA.

distribution

supposed distribution of the random variable. The default value is "lognormal". So far no other distribution is implemented.

seed

if seed = NULL a random seed is used. Otherwise the user can specify an integer for the seed.

details

a logical variable, where the default value is FALSE. If FALSE the output consists of a vector with simulated driving distances for the censored units regarding to their current operating time and the input driving distances for the failed units. If TRUE the output consists of a detailed list, i.e the same vector as described before, simulated annual driving distances, estimated distribution parameters and a seed for reproducibility.

Value

A numerical vector of simulated driving distances for the censored units and the input driving distances for the failed units if details = FALSE. If details = TRUE the output is a list which consists of the following elements:

• mileage : Simulated driving distances for the censored units and the input driving distances for the failed units.

• mileage_sim_annual : Simulated annual driving distances of specified distribution with estimated parameters. The length of x_sim is equal to the number of censored observations.

• coefficients : Estimated coefficients of supposed distribution.

• int_seed : Integer seed number for reproducibility.

Examples

# NOT RUN {
# Example 1 - Simplified vector output (complete data):
date_of_registration <- c("2014-08-17", "2014-03-29", "2014-12-06",
                          "2014-09-09", "2014-05-14", "2014-07-01",
                          "2014-06-16", "2014-04-03", "2014-05-23",
                          "2014-05-09", "2014-05-31", "2014-08-12",
                          "2014-04-13", "2014-02-15", "2014-07-07",
                          "2014-03-12", "2014-05-27", "2014-06-02",
                          "2014-05-20", "2014-03-21", "2014-06-19",
                          "2014-02-12", "2014-03-27")
date_of_repair <- c("2014-10-21", "2014-09-15", "2015-07-04", "2015-04-10",
                    "2015-02-15", "2015-04-14", "2015-04-24", "2015-02-27",
                    "2015-04-25", "2015-04-24", "2015-06-12", "2015-08-26",
                    "2015-05-04", "2015-04-04", "2015-09-06", "2015-05-22",
                    "2015-08-21", "2015-09-17", "2015-09-15", "2015-08-15",
                    "2015-11-26", "2015-08-22", "2015-10-05")

op_time <- as.numeric(difftime(as.Date(date_of_repair),
                               as.Date(date_of_registration),
                               units = "days"))
mileage <- c(5227, 15655, 13629, 18292, 24291, 34455, 33555, 21659, 21737,
             29870, 21068, 22986, 122283, 31592, 49050, 36088, 10918, 11153,
             122437, 122842, 20349, 65656, 40777)
state <- sample(c(0, 1), size = length(op_time), replace = TRUE)

mileage_corrected <- mcs_mileage(x = op_time, event = state,
                                 mileage = mileage,
                                 distribution = "lognormal", seed = NULL,
                                 details = FALSE)

# Example 2 - Detailed list output (complete data):
list_detail <- mcs_mileage(x = op_time, event = state, mileage = mileage,
                               distribution = "lognormal", seed = NULL,
                               details = TRUE)

# Example 3 - Detailed list output (realistic example):
op_time <- c(65, 170, 210, 213, 277, 287, 312, 330, 337, 350, 377, 379, 386,
             413, 426, 436, 451, 472, 483, 512, 525, 556, 557)
mileage <- c(NA, 15655, 13629, NA, 24291, 34455, NA, 21659, 21737,
             NA, 21068, 22986, NA, 31592, 49050, NA, 10918, 11153,
             NA, 122842, 20349, NA, 40777)
state <- c(0, 1, 1, 0, 1, 1, 0, 1, 1, 0, 1, 1, 0, 1, 1, 0, 1, 1, 0,
           1, 1, 0, 1)

list_detail <- mcs_mileage(x = op_time, event = state, mileage = mileage,
                               distribution = "lognormal", seed = NULL,
                               details = TRUE)
# }