2011/10/26 Jacob VanderPlas <[email protected]>:
> Olivier Grisel wrote:
>> A note for the scikit-learn developers:
>>
>> => we should definitely improve the tooling for checking the input and
>> emit informative ValueError messages that state explicitly that
>> scipy.sparse matrices are not supported as input for the models
>> mentioned by the poster.
>>
>
> Would it be worth creating an input-checking utility that would give
> unified output across scikit-learn?  I'm thinking something like
>
> utils.check_input(X, array=True, matrix=True, sparse=False)
>
> which would let a np array or matrix pass through, but raise an
> informative error for a sparse format.  We could extend this to other
> types, or even allow dimensionality checks as well.

larsmans already started to work on uniformizing those checks and
there is already:

https://github.com/scikit-learn/scikit-learn/blob/master/sklearn/utils/__init__.py#L60
https://github.com/scikit-learn/scikit-learn/blob/master/sklearn/utils/__init__.py#L65
https://github.com/scikit-learn/scikit-learn/blob/master/sklearn/utils/__init__.py#L93

However `array2d` should be improved to explicitly reject scipy.sparse
matrices with an explicit error message. And `check_arrays` parameter
`sparse_format` should be extended to add the value `invalid` (or
`reject` for instance) that raises a ValueError with the same explicit
message as well.

-- 
Olivier
http://twitter.com/ogrisel - http://github.com/ogrisel

------------------------------------------------------------------------------
The demand for IT networking professionals continues to grow, and the
demand for specialized networking skills is growing even more rapidly.
Take a complimentary Learning@Cisco Self-Assessment and learn 
about Cisco certifications, training, and career opportunities. 
http://p.sf.net/sfu/cisco-dev2dev
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general

Reply via email to