Thanks very much Sean, That is very quick answer! I use the second option, set fixed trees for each crossover pipeline as I think in this way there will be some time saved. If the tree is not fixed, two trees will be chosen randomly and they have to be the same(have same constrains)to be able to continue. Is this scenario odd? if the ADF does not appear in the main tree howeve the crossover has been done use the ADF tree therefor the individual will be the same before and after the crossover at least in terms of fitness function. Zheng