power_law_process: Maximum Likelihood Estimates for a Power Law Process

In jjw3952/mcotear: Contains helpful functions for use within MCOTEA

Maximum Likelihood Estimates for a Power Law Process

Description

power_law_process calculates the Maximum Likelihood Estimates for a Nonhomogeneous Poisson Process (NHPP) with Power Law Intensity Function (Crow-AMSAA model).

Usage

power_law_process(t, T, alpha = 0.05, fail.trunc = FALSE, iter = 10)

Arguments

t	A list of failure time vectors. Each vector should indicate a different system, i.e. if you have multiple systems each systems' failure times should be in it's own vector.
T	A list of Total Time on Test (TTT) (i.e. test duration) vectors. The vectors in the list should be of length 1, and each vector should indicate a different system, i.e. if you have multiple systems each systems' TTT should be in it's own vector.
alpha	1-confidence, supplied for parameter confidence intervals.
fail.trunc	Logical indicating if the test was failure terminated.
iter	The number of iterations for parameter calculations when fail.truc is TRUE.

Value

The output will be a list consisting of parameter estimates, confidence interals (exact confidence intervals for beta), and iterative beta and lambda estimates to verify convergence when fail.truc is TRUE.

See Also

power_law_mcf, mcf, trend_test, ttt, common_beta

Examples

data(amsaa)

# Three systems all time truncated at 200 hours
power_law_process(
 t = split(amsaa$Time, amsaa$System),
 T = list(200,200,200),
 alpha = 0.05,
 fail.trunc = FALSE,
 iter = 10
)

# Three systems all failure truncated
power_law_process(
  t = split(amsaa$Time, amsaa$System),
  T = list(197.2,190.8,195.8),
  alpha = 0.19,
  fail.trunc = TRUE,
  iter = 10
)

# One system time truncated at 200 hours
power_law_process(
  t = list(subset(amsaa, System == "S1")$Time),
  T = list(200),
  alpha = 0.05,
  fail.trunc = FALSE,
  iter = 10
)

rm(list = c("amsaa"))

jjw3952/mcotear documentation built on Sept. 2, 2023, 10:30 a.m.