predict {rEMM} | R Documentation |
Predict a Future State
Description
Predict a state or the probability distribution over states in n
time steps.
Usage
## S4 method for signature 'TRACDS'
predict(object, current_state = NULL, n=1,
probabilities = FALSE, randomized = FALSE, prior=FALSE)
Arguments
object |
an |
current_state |
use a specified current state.
If |
n |
number of time steps. |
probabilities |
if |
randomized |
if |
prior |
add one to each transition count. This is equal to starting with a uniform prior for the transition count distribution, i.e. initially all transitions are equally likely. It also prevents the product of probabilities to be zero if a transition was never observed. |
Details
Prediction is done using A^n
where A
is the transition
probability matrix maintained by the EMM.
Random tie-breaking is used.
Value
The name of the predicted state or a vector with the probability distribution over all states.
See Also
Examples
data("EMMTraffic")
emm <- EMM(measure="eJaccard", threshold=0.2)
emm <- build(emm, EMMTraffic)
#plot(emm) ## plot graph
## Predict state starting an state 1 after 1, 2 and 100 time intervals
## Note, state 7 is an absorbing state.
predict(emm, n=1, current_state="1")
predict(emm, n=2, current_state="1")
predict(emm, n=100, current_state="1")
## Get probability distribution
predict(emm, n=2, current_state="1", probabilities = TRUE)