em: EM step

Description Usage Arguments Details Value

Description

Execute one EM step for software reliability models.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
em_exp_emstep(params, data)

em_gamma_emstep(params, data, divide = 15L, eps = 1e-10)

em_llogis_emstep(params, data)

em_llogis_estep(params, data)

em_llogis_pllf(params, data, w1)

em_lnorm_emstep(params, data)

em_lxvmax_estep(params, data)

em_lxvmax_pllf(params, data, w1)

em_lxvmin_estep(params, data)

em_lxvmin_pllf(params, data, w1)

em_pareto_emstep(params, data)

em_tlogis_emstep_mo(params, data)

em_tlogis_estep(params, data)

em_tlogis_pllf(params, data, w0, w1)

em_tnorm_emstep(params, data)

em_txvmax_emstep_mo(params, data)

em_txvmax_estep(params, data)

em_txvmax_pllf(params, data, w0, w1)

em_txvmin_estep(params, data)

em_txvmin_pllf(params, data, w0, w1)

Arguments

params

A numeric vector for model parameters

data

A faultdata

divide

An integer for the number of integration points.

eps

A numeric value for tolerance error.

w1

A numeric value indicating the expected number of faults after te.

w0

A numeric value indicating the expected number of faults before 0.

...

Other parameters

Details

em_tlogis_emstep_mo is an emstep based on Marshall-Olkin-type (maximum) with Exp

em_tnorm_emstep has been modified by using EM for truncated distribution. Concretely, when the distribution is truncated at origin, the expected value is given by

\text{E}[h(X)|D] = ∑_{i=1}^k h(x_i) + \frac{k}{\overline{F}(0)} \int_{-∞}^0 h(x) f(x) dx.

em_txvmax_emstep_mo is an emstep based on Marshall-Olkin-type (maximum) with Exp

Value

A list with an updated parameter vector (param), absolute difference of parameter vector (pdiff), log-likelihood function for a given parameter vector (llf), the number of total faults (total).


okamumu/Rsrat documentation built on Oct. 3, 2018, 2:36 a.m.