genmarkov

genmarkov — generates random markov matrix with recurrent and transient classes

Calling sequence

M=genmarkov(rec,tr)  
M=genmarkov(rec,tr,flag)  

Parameters

rec : integer row vector (its dimension is the number of recurrent classes).
tr : integer (number of transient states)
M : real Markov matrix. Sum of entries in each row should add to one.
flag : string 'perm'. If given, a random permutation of the states is done.

Description

Returns in M a random Markov transition probability matrix with size(rec,1) recurrent classes with rec(1),...rec($) entries respectively and tr transient states.

Examples



//P has two recurrent classes (with 2 and 1 states) 2 transient states
P=genmarkov([2,1],2,'perm')
[perm,rec,tr,indsRec,indsT]=classmarkov(P);
P(perm,perm)
 
  

See also

classmarkov, eigenmarkov