Gunnar Raetsch benchmark datasets added

.travis.yml

@@ -16,10 +16,10 @@ install:
   - chmod +x ./cc-test-reporter
   - ./cc-test-reporter before-build
 script:
-  - nosetests -x -v -d --with-coverage --cover-erase --cover-package=skdatasets --cover-xml tests/*/*/test_abalone* tests/test_load.py
+  - nosetests -x -v -d --with-coverage --cover-erase --cover-package=skdatasets --cover-xml tests/*/*/test_abalone* tests/gunnar_raetsch/test_load.py tests/test_load.py
   - mkdir docs
   - export PYTHONPATH=`pwd`
-  - sphinx-quickstart -q -p scikit-datasets -a "David Diaz Vico" -v 0.1 -r 0.1.11 -l en --ext-autodoc --ext-viewcode --ext-githubpages --extensions sphinxcontrib.napoleon --no-makefile --no-batchfile docs
+  - sphinx-quickstart -q -p scikit-datasets -a "David Diaz Vico" -v 0.1 -r 0.1.12 -l en --ext-autodoc --ext-viewcode --ext-githubpages --extensions sphinxcontrib.napoleon --no-makefile --no-batchfile docs
   - sphinx-apidoc -o docs/_static/ skdatasets -F -a -l
   - travis-sphinx -v build -s docs -n
 after_success:

setup.py

@@ -3,12 +3,12 @@
 
 setup(name='scikit-datasets',
       packages=find_packages(),
-      version='0.1.11',
+      version='0.1.12',
       description='Scikit-learn-compatible datasets',
       author='David Diaz Vico',
       author_email='david.diaz.vico@outlook.com',
       url='https://github.com/daviddiazvico/scikit-datasets',
-      download_url='https://github.com/daviddiazvico/scikit-datasets/archive/v0.1.11.tar.gz',
+      download_url='https://github.com/daviddiazvico/scikit-datasets/archive/v0.1.12.tar.gz',
       keywords=['scikit-learn', 'pandas'],
       classifiers=['Intended Audience :: Science/Research',
                    'Intended Audience :: Developers',

skdatasets/__init__.py

@@ -15,6 +15,11 @@
                               load_digits, load_iris, load_linnerud, load_wine)
 from sklearn.model_selection import PredefinedSplit
 
+from .gunnar_raetsch import (load_banana, load_breast_cancer, load_diabetis,
+                             load_flare_solar, load_german, load_heart,
+                             load_image, load_ringnorm, load_splice,
+                             load_thyroid, load_titanic, load_twonorm,
+                             load_waveform)
 from .keel.imbalanced import (load_abalone9_18, load_abalone19,
                               load_cleveland_0_vs_4, load_ecoli4,
                               load_ecoli_0_1_3_7_vs_2_6,
@@ -59,6 +64,16 @@
 
 loader = dict()
 
+loader['gunnar_raetsch'] = {'banana': load_banana,
+                            'breast_cancer': load_breast_cancer,
+                            'diabetis': load_diabetis,
+                            'flare_solar': load_flare_solar,
+                            'german': load_german, 'heart': load_heart,
+                            'image': load_image, 'ringnorm': load_ringnorm,
+                            'splice': load_splice, 'thyroid': load_thyroid,
+                            'titanic': load_titanic, 'twonorm': load_twonorm,
+                            'waveform': load_waveform}
+
 
 loader['keel'] = {'abalone9-18': load_abalone9_18, 'abalone19': load_abalone19,
                   'cleveland-0_vs_4': load_cleveland_0_vs_4,
@@ -135,7 +150,13 @@
 
 def load(collection, name):
     """ Load a dataset. """
-    if collection == 'keel':
+    if collection == 'gunnar_raetsch':
+        data = loader['gunnar_raetsch'][name]()
+        X = data.features
+        y = data.target
+        X_test = y_test = inner_cv = None
+        outer_cv = data.splits
+    elif collection == 'keel':
         data = loader['keel'][name]()
         X = np.vstack(data.data5_test)
         y = np.hstack(data.target5_test)

skdatasets/gunnar_raetsch/__init__.py

@@ -0,0 +1,34 @@
+"""
+Gunnar Raetsch benchmark datasets
+(https://github.com/tdiethe/gunnar_raetsch_benchmark_datasets).
+
+@author: David Diaz Vico
+@license: MIT
+"""
+
+from functools import partial
+
+from .base import load_dataset
+
+
+load_banana = partial(load_dataset, name='banana')
+load_breast_cancer = partial(load_dataset, name='breast_cancer')
+load_diabetis = partial(load_dataset, name='diabetis')
+load_flare_solar = partial(load_dataset, name='flare_solar')
+load_german = partial(load_dataset, name='german')
+load_heart = partial(load_dataset, name='heart')
+load_image = partial(load_dataset, name='image')
+load_ringnorm = partial(load_dataset, name='ringnorm')
+load_splice = partial(load_dataset, name='splice')
+load_thyroid = partial(load_dataset, name='thyroid')
+load_titanic = partial(load_dataset, name='titanic')
+load_twonorm = partial(load_dataset, name='twonorm')
+load_waveform = partial(load_dataset, name='waveform')
+
+
+load = {'banana': load_banana, 'breast_cancer': load_breast_cancer,
+        'diabetis': load_diabetis, 'flare_solar': load_flare_solar,
+        'german': load_german, 'heart': load_heart, 'image': load_image,
+        'ringnorm': load_ringnorm, 'splice': load_splice,
+        'thyroid': load_thyroid, 'titanic': load_titanic,
+        'twonorm': load_twonorm, 'waveform': load_waveform}

skdatasets/gunnar_raetsch/base.py

@@ -0,0 +1,107 @@
+"""
+Gunnar Raetsch benchmark datasets
+(https://github.com/tdiethe/gunnar_raetsch_benchmark_datasets).
+
+@author: David Diaz Vico
+@license: MIT
+"""
+
+from scipy.io import loadmat
+from sklearn.model_selection import BaseCrossValidator
+
+from ..base import Bunch
+
+
+class GunnarRaetschDatasetSplit(BaseCrossValidator):
+    """Predefined split cross-validator for Gunnar Raetsch datasets.
+
+    Provides train/test indices to split data into train/test sets using a
+    predefined scheme.
+
+    Read more in the :ref:`User Guide <cross_validation>`.
+
+    Parameters
+    ----------
+    train_splits: array-like, shape (n_samples,)
+                  List of indices for each training split.
+    test_splits: array-like, shape (n_samples,)
+                 List of indices for each test split.
+
+    """
+
+    def __init__(self, train_splits, test_splits):
+        self.train_splits = train_splits - 1
+        self.test_splits = test_splits - 1
+
+    def split(self, X=None, y=None, groups=None):
+        """Generate indices to split data into training and test set.
+
+        Parameters
+        ----------
+        X: object
+            Always ignored, exists for compatibility.
+        y: object
+            Always ignored, exists for compatibility.
+        groups: object
+            Always ignored, exists for compatibility.
+
+        Returns
+        -------
+        train: ndarray
+            The training set indices for that split.
+        test: ndarray
+            The testing set indices for that split.
+
+        """
+        for train_indices, test_indices in zip(self.train_splits, self.test_splits):
+            yield (train_indices, test_indices)
+
+    def get_n_splits(self, X=None, y=None, groups=None):
+        """Returns the number of splitting iterations in the cross-validator
+
+        Parameters
+        ----------
+        X: object
+           Always ignored, exists for compatibility.
+        y: object
+           Always ignored, exists for compatibility.
+        groups: object
+                Always ignored, exists for compatibility.
+
+        Returns
+        -------
+        n_splits: int
+                  Returns the number of splitting iterations in the
+                  cross-validator.
+
+        """
+        return len(self.train_splits)
+
+
+def load_dataset(name, return_X_y=False):
+    """Load dataset.
+
+    Load a dataset.
+
+    Parameters
+    ----------
+    name: string
+          Dataset name.
+    return_X_y: bool, default=False
+                If True, returns (data, target) instead of a Bunch object..
+
+    Returns
+    -------
+    data: Bunch
+          Dictionary-like object with all the data and metadata.
+    X, y: arrays
+          If return_X_y is True
+
+    """
+    features, target, train_splits, test_splits = loadmat('skdatasets/gunnar_raetsch/benchmarks')[name][0][0]
+
+    if return_X_y:
+        return features, target
+
+    return Bunch(features=features, target=target,
+                 splits=GunnarRaetschDatasetSplit(train_splits, test_splits))

tests/gunnar_raetsch/test_load.py

@@ -0,0 +1,61 @@
+"""
+Tests.
+
+@author: David Diaz Vico
+@license: MIT
+"""
+
+from sklearn.linear_model import LogisticRegression
+from sklearn.model_selection import cross_val_score
+
+from ..base import check_load_dataset
+
+from skdatasets.gunnar_raetsch import (load_banana, load_breast_cancer,
+                                       load_diabetis, load_flare_solar,
+                                       load_german, load_heart, load_image,
+                                       load_ringnorm, load_splice, load_thyroid,
+                                       load_titanic, load_twonorm,
+                                       load_waveform)
+
+
+def test_split():
+    """Tests Gunnar Raetsch dataset splits."""
+    data = load_banana()
+    X = data.features
+    y = data.target
+    splits = data.splits
+    cross_val_score(LogisticRegression(), X, y=y, cv=splits)
+
+
+def test_load():
+    """Tests Gunnar Raetsch benchmark datasets."""
+    datasets = {'banana': {'loader': load_banana, 'n_patterns': (5300, ),
+                           'n_variables': 2},
+                'breast_cancer': {'loader': load_breast_cancer,
+                                  'n_patterns': (263, ), 'n_variables': 9},
+                'diabetis': {'loader': load_diabetis, 'n_patterns': (768, ),
+                             'n_variables': 8},
+                'flare_solar': {'loader': load_flare_solar,
+                                'n_patterns': (144, ), 'n_variables': 9},
+                'german': {'loader': load_german, 'n_patterns': (1000, ),
+                           'n_variables': 20},
+                'heart': {'loader': load_heart, 'n_patterns': (270, ),
+                          'n_variables': 13},
+                'image': {'loader': load_image, 'n_patterns': (2086, ),
+                          'n_variables': 18},
+                'ringnorm': {'loader': load_ringnorm, 'n_patterns': (7400, ),
+                             'n_variables': 20},
+                'splice': {'loader': load_splice, 'n_patterns': (2991, ),
+                           'n_variables': 60},
+                'thyroid': {'loader': load_thyroid, 'n_patterns': (215, ),
+                            'n_variables': 5},
+                'titanic': {'loader': load_titanic, 'n_patterns': (24, ),
+                            'n_variables': 3},
+                'twonorm': {'loader': load_twonorm, 'n_patterns': (7400, ),
+                            'n_variables': 20},
+                'waveform': {'loader': load_waveform, 'n_patterns': (5000, ),
+                             'n_variables': 21}}
+    for dataset in datasets.values():
+        check_load_dataset(dataset['loader'], dataset['n_patterns'],
+                           dataset['n_variables'], (('features', 'target'), ),
+                           n_targets=1, n_folds=None)

tests/test_load.py

@@ -12,6 +12,8 @@
 
 def test_load():
     """Tests load."""
+    X, y, X_test, y_test, inner_cv, outer_cv = load('gunnar_raetsch', 'banana')
+    check_items([X, y, outer_cv], [X_test, y_test, inner_cv])
     X, y, X_test, y_test, inner_cv, outer_cv = load('keel', 'abalone9-18')
     check_items([X, y, outer_cv], [X_test, y_test, inner_cv])
     try: