Missing 'query_id' -- breaks API compatibility with sklearn

[h4ste]

Source: http://scikit-learn.org/stable/modules/generated/sklearn.datasets.load_svmlight_file.html

sklearn.datasets.load_svmlight_file(f, n_features=None, dtype=<type 'numpy.float64'>, multilabel=False, zero_based='auto', query_id=False)

This implementation doesn't support the query_id argument, which breaks drop-in API compatibility.

[mblondel]

Indeed, adding support for query_id would be nice. I just never had the need for it yet.

The loader in scikit-learn also supports reading from file objects, so this implementation will never be an exact drop-in replacement.