The documentation is pretty decent. You could add something like 
"trust us it works"... ?

But anyway, I don't trust anything my computer does, and habitually
reach for the print statement so I can check that things are not
going off-the-rails. As they do often.

I'm mightily impressed by the coverage of sklearn, and understand
a bit the tradeoffs that one must make to get all this functionality.
Having said that, I have often wanted to "verbosify" the internals,
so I can see how the learning procedures are going. I already hacked
apart the Boosting code (quite easy) and now I ended up instrumenting
the tree code (a bit harder) to find out what was happening.

BTW, the code documents itself pretty well, so I wouldn't bother trying to
document the internals.

Simon.

On Sat, 15 Aug 2015 23:53:24 +1000
Joel Nothman <joel.noth...@gmail.com> wrote:

> While it's not bad to have more people know the internals of the tree code,
> ideally people shouldn't *have* to. Do you have any hints for how
> documentation could better serve users to not land in whatever trap you did?
> 
> On 15 August 2015 at 16:03, Simon Burton <si...@arrowtheory.com> wrote:
> 
> >
> > My bad. I did something stupid (again).
> >
> > On the plus side, I now know my way around the internals of
> > the tree code much better.
> >
> > Cheers.
> >
> >
> > On Sat, 15 Aug 2015 14:11:49 +1000
> > Simon Burton <si...@arrowtheory.com> wrote:
> >
> > >
> > > Hi,
> > >
> > > I am training a DecisionTreeClassifier on samples with a large (500)
> > > number of features. I find that the tree refuses to grow (and so
> > > cannot be used in boosting) unless I remove (zero) some of the
> > > features. This seems strange. Any ideas why? I tried fiddling
> > > with the settings, now delving into the implementation.
> > >
> > > Simon.
> > >
> > >
> > ------------------------------------------------------------------------------
> > > _______________________________________________
> > > Scikit-learn-general mailing list
> > > Scikit-learn-general@lists.sourceforge.net
> > > https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
> >
> >
> > ------------------------------------------------------------------------------
> > _______________________________________________
> > Scikit-learn-general mailing list
> > Scikit-learn-general@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
> >

------------------------------------------------------------------------------
_______________________________________________
Scikit-learn-general mailing list
Scikit-learn-general@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general

Reply via email to