Hi Miquel, These options are not available within RandomForestClassifier/Regressor. By default len(X) are drawn with replacement.
However, you can achieve what you look for using BaggingClassifier(base_estimator=DecisionTreeClassifier(...), max_samples=..., max_features=...), where max_samples and max_features are the number of samples/features drawn to be build each base estimator. If you want to draw samples (resp features) with replacement, set bootstrap=True (resp bootstrap_features=True). Best, Gilles On 16 December 2014 at 09:06, Miquel Camprodon <miquel.campro...@kernel-analytics.com> wrote: > Hi all, > > I am a newbie in the use of scikit-learn. Congratulations for the project! > > (1) I am using the ensemble methods RandomForestRegressor and > RandomForestClassifier. I would like to adjust the number of subsamples each > estimator will use. How can I achieve this? My background is in R, and in > that language there is the sampsize parameter which allows to control it. > If it is not possible, how are subsamples drawn for each estimator in > scikit-learn? > > (2) How can I parametrize the number of features each estimator will use? > > (3) If the answer is that I should write my own ensemble, can you give me > some hints about how to do it? > > Thanks a lot, > > Miquel > > Aviso legal: Este mensaje, incluidos sus ficheros adjuntos, contienen > información confidencial, dirigida a un destinatario y con una finalidad > específica protegida por la Ley. Le informamos que si Usted no es el > destinatario del mismo, deberá eliminar éste mensaje, quedando estrictamente > prohibida cualquier divulgación, copia o distribución de su contenido, así > como desarrollar o ejecutar cualquier tipo de acción basada en su contenido. > Si Usted es el destinatario, el uso de la información que ha recibido, > deberá estar limitado a la finalidad específica por la cual se le ha > enviado. > > Disclaimer: This message, including its attachments, contains confidential > information addressed to a recipient for a specific purpose protected by > law. Please note that if you are not the intended addressee of this message > you must immediately delete the message and its attachments. Any disclosure, > copying or distribution of the contents of this message is strictly > forbidden. The execution of any action based on the contents of this message > is strictly forbidden. If you are the intended recipient, the use of the > information received shall be limited to the specific purpose for which it > has been sent. > > Avís legal: Aquest missatge, inclosos els seus fitxers adjunts, conté > informació confidencial, dirigida a un destinatari i amb una finalitat > específica protegida per la Llei. L'informem que si Vostè no és el > destinatari del mateix, haurà d’eliminar aquest correu, quedant estrictament > prohibida qualsevol divulgació, còpia o distribució del seu contingut, així > com desenvolupar o executar qualsevol tipus d’acció basada en el seu > contingut. Si Vostè és el destinatari, l’ús de la informació que ha rebut > haurà d’estar limitat a la finalitat específica per la qual se li ha enviat. > > ------------------------------------------------------------------------------ > Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server > from Actuate! Instantly Supercharge Your Business Reports and Dashboards > with Interactivity, Sharing, Native Excel Exports, App Integration & more > Get technology previously reserved for billion-dollar corporations, FREE > http://pubads.g.doubleclick.net/gampad/clk?id=164703151&iu=/4140/ostg.clktrk > _______________________________________________ > Scikit-learn-general mailing list > Scikit-learn-general@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/scikit-learn-general > ------------------------------------------------------------------------------ Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server from Actuate! Instantly Supercharge Your Business Reports and Dashboards with Interactivity, Sharing, Native Excel Exports, App Integration & more Get technology previously reserved for billion-dollar corporations, FREE http://pubads.g.doubleclick.net/gampad/clk?id=164703151&iu=/4140/ostg.clktrk _______________________________________________ Scikit-learn-general mailing list Scikit-learn-general@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
