Funtion to merge states based on their state means
mergeFunc takes the output of hmm.run.func (or find.hmm.states)
with a particular model selection criterion and iteratively merges the
states with means closer than a supplied threshold.
mergeHmmStates is a frontend for
mergeHmmStates(aCGH.obj, model.use = 1, minDiff = 0.25) mergeFunc(statesres, minDiff = 0.1)
Object of class
- the sublist of the states.hmm list output from
find.hmm.statesfor a given model selection crterion
- The states whose predicted values are less than minDiff apart are merged into one state and all the predicited values are recomputed.
Model selection criterion to use, See
This function is intended to reduce effect of the possible small magnitude technological artifacts on the structure determination.
List containing states.hmm object is returned.
Application of Hidden Markov Models to the analysis of the array CGH data, Fridlyand et.al., JMVA, 2004