[email protected]>"
mailto:[email protected]>>
Fecha: jueves, 3 de julio de 2014 12:51
Para: scikit-learn-general
mailto:[email protected]>>
Asunto: Re: [Scikit-learn-general] Concatenating scikit.sparse matrix and nump
s.sourceforge.net>
> Fecha: jueves, 3 de julio de 2014 08:39
> Para: scikit-learn-general
> Asunto: Re: [Scikit-learn-general] Concatenating scikit.sparse matrix and
> numpy arrays
>
> I think you may be reinventing sklearn.pipeline.FeatureUnion. If one of
> the transfo
[email protected]>"
mailto:[email protected]>>
Fecha: jueves, 3 de julio de 2014 08:39
Para: scikit-learn-general
mailto:[email protected]>>
Asunto: Re: [Scikit-learn-general] Concatenating scikit.sparse matrix and numpy
arra
I think you may be reinventing sklearn.pipeline.FeatureUnion. If one of the
transformers returns sparse, it will hstack all the outputs to a sparse
format.
On 3 July 2014 02:35, ZORAIDA HIDALGO SANCHEZ <
[email protected]> wrote:
> Dear all,
>
> For a given dataset, I can hav
Dear all,
For a given dataset, I can have more than one source(differente csv) and
each of these sources needs a different transformer(for instance, one
source could be text and
uses TfidfVectorizer whereas other is composed by attributes of type float
and needs to be normalized using StandardScal
