| markov.disturbance {spgs} | R Documentation |
Construct feasible Random Noise Generating a First-Order Markov Chain
Description
Produces a sequence of random noise which would generate an observed sequence of finite symbols provided that the sequence of symbols results from a first-order Markov chain.
Usage
markov.disturbance(x, chain = NULL, random = TRUE, bandwidth = 1,
estimates = is.null(chain))
Arguments
x |
A sequence of finite symbols represented as a character vector. |
chain |
A list containing two named components which specify a first-order Markov chain.
The ‘trans.mat’ component holds the stochastic transition matrix for the
chain while the ‘stat.dist’ component holds the stationary distribution of
the chain. If not specified, ‘chain’ is estimated from ‘x’ using
|
random |
This can be a logical value or a number in the range 0-1. If ‘TRUE’, random noise will be generated. If ‘FALSE’, the constant value 0.5 will be used as the noise source. If a value in the range 0-1 is specified, that value will be used as a constant noise source. the default value is ‘TRUE’. |
bandwidth |
This value, which should be in the range 0-1, specifies the maximum peak-to-peak bandwidth of the random noise generated. The default value is 1. |
estimates |
A logical value specifying if the Markov chain estimates should be included in the return. |
Value
If ‘estimates’ is ‘TRUE’, returns a list containing the following components:
disturbance |
the sequence of random noise as a numeric vector. |
trans.mat |
The stochastic transition matrix estimated from x, if ‘chain’ is NULL; otherwise a copy of ‘trans.mat’ component of ‘chain’. |
stat.dist |
The stationary distribution estimated from x, if ‘chain’ is NULL; otherwise a copy of the ‘stat.dist’ component of ‘chain’. |
Otherwise, if ‘estimate’ is ‘FALSE’, returns the sequence of random noise as a numeric vector.
Author(s)
Andrew Hart and Servet MartÃnez
See Also
markov.test, diid.test, diid.disturbance