The LDA accronym for Linear Discriminant Analysis is ambiguous
because of the newly introduced Latent Dirichlet Allocation model.
We therefore deprecate the sklearn.lda.LDA and sklearn.lda.QDA
in favor of explicit names.
Passing 1D arrays to check_array, without setting `ensure_2d` to false now
raises a deprecation warning before reshaping it. This will later throw an
error.
All Scaler classes also throw warnings when 1D arrays are passed.
All unit tests/doctests are modified to ensure that no 1D arrays are passed,
except in explicit 1D array tests where the warnings have been silenced.
Additional tests are also included which check for different 1D array cases.
2D array tests with one samples and one features are also added and where
they failed, `check_array` call has been modified to give a more useful error
message
Changed SubjectIndependentKFold to DisjointGroupKFold
cosmetic changes test (fix seed correctly, use assert_equal for
meaningful error messages)
Changed name to DisjointLabelKFold
Added example of use
FIX: whitespace related doctest failure
FIX: Python 2.6 requires the field numbers in print
FIX: change docstring to comment in test function
DOC: moved docstring from function to class
FIX: added call to parent class
FIX: error in calling the parent
DOC: fixed doctest
FIX: doctest
Cosmetic changes (minor refactoring)
Optimised code (use np.bincount)
Cosmetic: use samples instead of weight for clarity
Minor fix: removed shuffle parameter
Cosmetic
Use mergesort instead of quicksort for reproducibility.
Changed variable name 'y' to 'label'.
Added test for degenerate case where n_folds > n_labels.
Documented the requirement n_labels > n_folds.
DOC: improved description + added see also sections.
Fixed dtype of temporary arrays.
Improved test: check that one label is not in both test and training.
Added documentation for DisjoinLabelKFold.
This example trains several tree based ensemble methods and uses
them to transform the data into a high dimensional, sparse space.
The trains a linear model on this new feature space. The idea is
taken from:
Practical Lessons from Predicting Clicks on Ads at Facebook Junfeng Pan,
He Xinran, Ou Jin, Tianbing XU, Bo Liu, Tao Xu, Yanxin Shi, Antoine
Atallah, Ralf Herbrich, Stuart Bowers, Joaquin Quiñonero Candela
International Workshop on Data Mining for Online Advertising (ADKDD)
https://www.facebook.com/publications/329190253909587/
MAINT Remove sequence of sequence support from datasets
MAINT Remove return_indicator param
MAINT Remove multilabel-seq test in OVR
MAINT Remove multilable-seq test in check_cv
MAINT Remove multilabel seq test in label_binarizer
TST type_of_target returns "unknown" for multilabel-sequence types
TST _check_targets should raise a ValueError
DOC show multilabel indicator as an example; remove return_indicator param
DOC use consistent lower case y for target
* copyedit doc
* give the module an underscore (we should do this to all private modules...)
* use public import path in test
* change all logs to log1p to prevent warning in test and because it's
genuinely useful for frequency data
* what's new
Since `make_union` is a convenience factory method that returns a `FeatureUnion` class with names automatically filled in and obviously does *not* require manual naming of the components.
Closes#4724. The multi-class page now includes a list of estimators
that support multi-labels. Also adjusted the reference to LDA to include
QDA as a multiclass-supporting classifier.
Vinayak Mehta