SPORF
SPORF copied to clipboard
neurodata
Rerf can work for continuous y
say [x,y] from joint normal, then do RerF(x,y,trees=500), it outputs normally.
Jovo says it should not output unless y is categorical.
@MrAE @falkben @ttomita @jbrowne6 should lumberjack work currently if y is continuous?
Y should be modifiable to a factor. So Y could be an integer but each unique value will be considered a class.
James.
Get Outlook for iOShttps://aka.ms/o0ukef
From: joshua vogelstein [email protected] Sent: Saturday, October 20, 2018 7:32:13 PM To: neurodata/lumberjack Cc: James Browne; Mention Subject: Re: [neurodata/lumberjack] Rerf can work for continuous y (#71)
@MrAEhttps://github.com/MrAE @falkbenhttps://github.com/falkben @ttomitahttps://github.com/ttomita @jbrowne6https://github.com/jbrowne6 should lumberjack work currently if y is continuous?
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/neurodata/lumberjack/issues/71#issuecomment-431626082, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AVRR8V2jwz-aU2iXrLqYMQ50IVU3I-REks5um7J9gaJpZM4Xxd31.
A quick caveat:
The version of RerF with regression implemented will take a continuous Y but I’m not sure if that’s available yet.
James.
Get Outlook for iOShttps://aka.ms/o0ukef
From: James Browne Sent: Saturday, October 20, 2018 7:35:52 PM To: neurodata/lumberjack; neurodata/lumberjack Cc: Mention Subject: Re: [neurodata/lumberjack] Rerf can work for continuous y (#71)
Y should be modifiable to a factor. So Y could be an integer but each unique value will be considered a class.
James.
Get Outlook for iOShttps://aka.ms/o0ukef
From: joshua vogelstein [email protected] Sent: Saturday, October 20, 2018 7:32:13 PM To: neurodata/lumberjack Cc: James Browne; Mention Subject: Re: [neurodata/lumberjack] Rerf can work for continuous y (#71)
@MrAEhttps://github.com/MrAE @falkbenhttps://github.com/falkben @ttomitahttps://github.com/ttomita @jbrowne6https://github.com/jbrowne6 should lumberjack work currently if y is continuous?
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHubhttps://github.com/neurodata/lumberjack/issues/71#issuecomment-431626082, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AVRR8V2jwz-aU2iXrLqYMQ50IVU3I-REks5um7J9gaJpZM4Xxd31.
@jbrowne6 i don't understand your answer. Y is continuous, given that as input, i'd think it would spit out that it doesn't support regression yet (until regression is implemented). it seems it is not doing that?
@ttomita i think you implemented this, so you can make a PR for this, in addition to your smerf PR?