Great. Thank you for all your help. Cheers, Sarah
On Mon, Feb 5, 2018 at 12:56 AM, Joel Nothman <joel.noth...@gmail.com> wrote: > If you specify n_values=[list_of_vals_for_column1, > list_of_vals_for_column2], you should be able to engineer it to how you > want. > > On 5 February 2018 at 16:31, Sarah Wait Zaranek <sarah.zara...@gmail.com> > wrote: > >> If I use the n+1 approach, then I get the correct matrix, except with the >> columns of zeros: >> >> >>> test >> array([[0., 0., 0., 0., 0., 0., 0., 1., 1., 0., 0., 0., 0., 0., 1.], >> [0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 1., 1., 0., 0., 0.], >> [1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 1., 0., 0.], >> [0., 1., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 1., 0.]]) >> >> >> On Mon, Feb 5, 2018 at 12:25 AM, Sarah Wait Zaranek < >> sarah.zara...@gmail.com> wrote: >> >>> Hi Joel - >>> >>> Conceptually, that makes sense. But when I assign n_values, I can't >>> make it match the result when you don't specify them. See below. I used >>> the number of unique levels per column. >>> >>> >>> enc = OneHotEncoder(sparse=False) >>> >>> test = enc.fit_transform([[7, 0, 3], [1, 2, 0], [0, 2, 1], [1, 0, >>> 2]]) >>> >>> test >>> array([[0., 0., 1., 1., 0., 0., 0., 0., 1.], >>> [0., 1., 0., 0., 1., 1., 0., 0., 0.], >>> [1., 0., 0., 0., 1., 0., 1., 0., 0.], >>> [0., 1., 0., 1., 0., 0., 0., 1., 0.]]) >>> >>> enc = OneHotEncoder(sparse=False,n_values=[3,2,4]) >>> >>> test = enc.fit_transform([[7, 0, 3], [1, 2, 0], [0, 2, 1], [1, 0, >>> 2]]) >>> >>> test >>> array([[0., 0., 0., 1., 0., 0., 0., 1., 1.], >>> [0., 1., 0., 0., 0., 2., 0., 0., 0.], >>> [1., 0., 0., 0., 0., 1., 1., 0., 0.], >>> [0., 1., 0., 1., 0., 0., 0., 1., 0.]]) >>> >>> Cheers, >>> Sarah >>> >>> Cheers, >>> Sarah >>> >>> On Mon, Feb 5, 2018 at 12:02 AM, Joel Nothman <joel.noth...@gmail.com> >>> wrote: >>> >>>> If each input column is encoded as a value from 0 to the (number of >>>> possible values for that column - 1) then n_values for that column should >>>> be the highest value + 1, which is also the number of levels per column. >>>> Does that make sense? >>>> >>>> Actually, I've realised there's a somewhat slow and unnecessary bit of >>>> code in the one-hot encoder: where the COO matrix is converted to CSR. I >>>> suspect this was done because most of our ML algorithms perform better on >>>> CSR, or else to maintain backwards compatibility with an earlier >>>> implementation. >>>> >>>> _______________________________________________ >>>> scikit-learn mailing list >>>> scikit-learn@python.org >>>> https://mail.python.org/mailman/listinfo/scikit-learn >>>> >>>> >>> >> >> _______________________________________________ >> scikit-learn mailing list >> scikit-learn@python.org >> https://mail.python.org/mailman/listinfo/scikit-learn >> >> > > _______________________________________________ > scikit-learn mailing list > scikit-learn@python.org > https://mail.python.org/mailman/listinfo/scikit-learn > >
_______________________________________________ scikit-learn mailing list scikit-learn@python.org https://mail.python.org/mailman/listinfo/scikit-learn
