Re: [Scikit-learn-general] Number of subsamples in Random Forest

Hi Sebastian, Indeed, N samples are drawn with replacement, where N=len(original training set). I guess we could add an extra max_samples parameter, just like we have for the Bagging estimators. Gilles On 6 July 2015 at 23:00, Sebastian Raschka se.rasc...@gmail.com wrote: Thanks, Jeff, that

Re: [Scikit-learn-general] Number of subsamples in Random Forest

Thanks, Gilles, I think that's a good idea! It would make the implementation more flexible and would add clarity as well! Best, Sebastian On Jul 9, 2015, at 2:35 PM, Gilles Louppe g.lou...@gmail.com wrote: Hi Sebastian, Indeed, N samples are drawn with replacement, where N=len(original

Re: [Scikit-learn-general] Number of subsamples in Random Forest

Thanks, Jeff, that was what I expected, but I just wanted to make sure. I will add a note via a pull request later. On Jul 6, 2015, at 3:52 PM, Jeff Blackburne jblackbu...@gmail.com wrote: Hi Sebastian, The bootstrap sample size is always the same as the input sample size. If you feel

[Scikit-learn-general] Number of subsamples in Random Forest

Hi, all, I am wondering what the sample size of the bootstrap samples are in the RandomForestClassifier/Regressor (if bootstrap=True, default setting). From the documentation

Re: [Scikit-learn-general] Number of subsamples in Random Forest

Hi Sebastian, The bootstrap sample size is always the same as the input sample size. If you feel up to it, a pull request updating the documentation would probably be quite welcome. -Jeff On Mon, Jul 6, 2015 at 12:21 PM, Sebastian Raschka se.rasc...@gmail.com wrote: Hi, all, I am wondering