randomForestSRC individual tree predictions

[MNRiverEcologyUnit]

How can the individual tree predictions, OOB and or inbag, be accessed?

[ehrlinger]

Individual tree predictions are not especially useful, as the predicted value is the ensemble of all trees. You can get the OOB or inbag ensemble predictions using the predict.rfsrc function.

…

On Fri, Dec 30, 2016 at 1:47 PM, dtoalm ***@***.***> wrote: How can the individual tree predictions, OOB and or inbag, be accessed? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#1>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABDylIMSfEmYfA-2NZH6bCjXU-v1eaM1ks5rNVG4gaJpZM4LYVWM> .

[MNRiverEcologyUnit]

Thanks for getting back so quick. I’m new to the random forest technique and was trying to think of a way to provide some error around the prediction. I was thinking I could use the individual predictions to get a nonparametric confidence interval, but perhaps using the sqrt of the error rate would be a better option. Thanks again. Dan From: John Ehrlinger [mailto:notifications@github.com] Sent: Friday, December 30, 2016 1:58 PM To: ehrlinger/randomForestSRC <randomForestSRC@noreply.github.com> Cc: O'Shea, Daniel T (DNR) <Daniel.OShea@state.mn.us>; Author <author@noreply.github.com> Subject: Re: [ehrlinger/randomForestSRC] randomForestSRC individual tree predictions (#1) Individual tree predictions are not especially useful, as the predicted value is the ensemble of all trees. You can get the OOB or inbag ensemble predictions using the predict.rfsrc function.

On Fri, Dec 30, 2016 at 1:47 PM, dtoalm ***@***.******@***.***>> wrote: How can the individual tree predictions, OOB and or inbag, be accessed? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#1>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABDylIMSfEmYfA-2NZH6bCjXU-v1eaM1ks5rNVG4gaJpZM4LYVWM> .

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub<#1 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AK-R0GAGRCz0mKrq7yMP6gSA55A2wXUBks5rNWI5gaJpZM4LYVWM>.