md_score_weibull_series_C1_C2_C3: Score function generator (gradient of the log-likelihood) for...

In queelius/masked.data: Estimating Reliability of Components in Series from Masked Data with Relaxed Candidate Set Models

Score function generator (gradient of the log-likelihood) for Weibull series system on masked data, where the masked data is in the form of right-censored system lifetimes and masked component cause of failure. Note that if the right censoring indicator variable delta is missing, we assume that all system lifetimes are observed.

Description

Masked component data md approximately satisfies the following conditions: C1: Pr(K in C) = 1 C2: Pr(C=c | K=j, T=t) = Pr(C=c | K=j', T=t) for any j, j' in c. C3: masking probabilities are independent of theta

Usage

md_score_weibull_series_C1_C2_C3(
  md,
  sysvar = "t",
  setvar = "x",
  deltavar = "delta"
)

Arguments

md	masked data
sysvar	name of the system lifetime column
setvar	prefix symbol for the candidate sets column
deltavar	right-censoring indicator column name, if TRUE, then the system lifetime is right-censored (not observed), otherwise it is observed (not right-censored).

Value

a score function with respect to theta given md

queelius/masked.data documentation built on Jan. 28, 2025, 4:23 a.m.