mc: Create a Markov Chain Model

In yuki0425/mc: Markov Chain Model

Description

Create a Markov Chain Class.

Usage

mc(timecategory = "discrete", states, pijdef, qidef = NULL, name, byrow = TRUE)

dtmc(timecategory = 'discrete', states, pijdef, qidef = NULL, name, byrow = TRUE)

ctmc(timecategory = 'continuous', states, pijdef, qidef, name, byrow = TRUE)

Arguments

timecategory	indicate whether the Markov Chain is discrete time or continous time, default value is "discrete"
states	the states name of the Markov Chain
pijdef	the transition matrix of the Markov Chain
qidef	the holding-time rates for the continuous Markov Chain
name	the name of Markov Chain
byrow	logical, indicate whether transition probabilities are shown by row or by column, default value is TURE

Details

pijdef can either be in the matrix form or in the function form.

Author(s)

Jiexuan Cao, Yuki Liu, Qianhan Zhang

Examples

## discrete finite Markov Chain
dtftmc = mc(states = c("sunny", "cloudy", "rain"),
                 pijdef = function(i,j,...) {
                           if (i == 1 && j < 3 ||
                               i == 2 && j != 2 ||
                               i == 3 && j > 1)  return(0.5)
                           0
                        },  
       name = "Weather")

## discrete infinite Markov Chain
pijdeffunction = function(i,j) {
    p = 0.3
    q = 0.7
    r = 0
    if (i == 1 && j == 1)
        return (1 - p)
    if(j == i + 1)
        return(p)
    if(j == i - 1)
        return(q)
    if(j==i)
        return(r)
    return(0)
}

dtiftmc = mc(states = "infinity",
             pijdef = pijdeffunction, 
             name = "death and birth")
             
## continuous finite Markov Chain
ctftmc = mc(states = c("0", "1", "2"),
            pijdef = matrix(c(0, 0.2857143, 0, 1, 0, 1, 0, 0.7142857, 0), ncol = 3),
            qidef = c(0.25, 0.175, 0.08),
            name = 'Machine Repair')

yuki0425/mc documentation built on May 4, 2019, 7:44 p.m.