Re: [Scikit-learn-general] HMM: Monte Carlo Generation and Randon Number

2012-10-18 Thread Gael Varoquaux
On Thu, Oct 18, 2012 at 06:24:00PM +0100, Didier Vila wrote: > I would like to apply the same draw to HMM different models: is it possible ? Try setting 'random_state=0' in the constructor for both. That said, given that they are fit on different dataset, I am not sure what you can conclude from

[Scikit-learn-general] HMM: Monte Carlo Generation and Randon Number

2012-10-18 Thread Didier Vila
All, I am running Monte Carlo simulation on HMM. I would like to apply the same draw to HMM different models: is it possible ? Simple example: modelA=hmm.GaussianHMM(2) modelB=hmm.GaussianHMM(2) modelA.fit([dataA]) modelB.fit([dataB]) #first simulat