Finish to update the doc

Author: Guillaume Lemaitre

doc/api.rst

@@ -15,20 +15,21 @@ Under-sampling methods
 
 Classes
 -------
+.. currentmodule:: unbalanced_dataset
 
 .. autosummary::
    :toctree: generated/
 
-   unbalanced_dataset.under_sampling.ClusterCentroids
-   unbalanced_dataset.under_sampling.CondensedNearestNeighbour
-   unbalanced_dataset.under_sampling.EditedNearestNeighbours
-   unbalanced_dataset.under_sampling.RepeatedEditedNearestNeighbours
-   unbalanced_dataset.under_sampling.InstanceHardnessThreshold
-   unbalanced_dataset.under_sampling.NearMiss
-   unbalanced_dataset.under_sampling.NeighbourhoodCleaningRule
-   unbalanced_dataset.under_sampling.OneSidedSelection
-   unbalanced_dataset.under_sampling.RandomUnderSampler
-   unbalanced_dataset.under_sampling.TomekLinks
+   under_sampling.ClusterCentroids
+   under_sampling.CondensedNearestNeighbour
+   under_sampling.EditedNearestNeighbours
+   under_sampling.RepeatedEditedNearestNeighbours
+   under_sampling.InstanceHardnessThreshold
+   under_sampling.NearMiss
+   under_sampling.NeighbourhoodCleaningRule
+   under_sampling.OneSidedSelection
+   under_sampling.RandomUnderSampler
+   under_sampling.TomekLinks
 
 
 .. _over_sampling_ref:
@@ -42,12 +43,13 @@ Over-sampling methods
 
 Classes
 -------
+.. currentmodule:: unbalanced_dataset
 
 .. autosummary::
    :toctree: generated/
 
-   unbalanced_dataset.over_sampling.RandomOverSampler
-   unbalanced_dataset.over_sampling.SMOTE
+   over_sampling.RandomOverSampler
+   over_sampling.SMOTE
 
 
 .. _combine_ref:
@@ -61,12 +63,13 @@ Combination of over- and under-sampling methods
 
 Classes
 -------
+.. currentmodule:: unbalanced_dataset
 
 .. autosummary::
    :toctree: generated/
 
-   unbalanced_dataset.combine.SMOTEENN
-   unbalanced_dataset.combine.SMOTETomek
+   combine.SMOTEENN
+   combine.SMOTETomek
 
 .. _ensemble_ref:
 
@@ -79,9 +82,10 @@ Ensemble methods
 
 Classes
 -------
+.. currentmodule:: unbalanced_dataset
 
 .. autosummary::
    :toctree: generated/
 
-   unbalanced_dataset.ensemble.BalanceCascade
-   unbalanced_dataset.ensemble.EasyEnsemble
+   ensemble.BalanceCascade
+   ensemble.EasyEnsemble

examples/combine/README.txt

@@ -1,6 +1,6 @@
 .. _combine_examples:
 
-Example using combine class methods
-===================================
+Examples using combine class methods
+====================================
 
-Combine examples.
+Combine methods mixed over- and under-sampling methods. Generally SMOTE is used for over-sampling while some cleaning methods (i.e., ENN and Tomek links) are used to under-sample.

examples/combine/plot_smote_enn.py

@@ -3,7 +3,7 @@
 SMOTE + ENN
 ===========
 
-An illustration of the random SMOTE + ENN method.
+An illustration of the SMOTE + ENN method.
 
 """
 
@@ -33,9 +33,9 @@
 # Fit and transform x to visualise inside a 2D feature space
 X_vis = pca.fit_transform(X)
 
-# Apply the random under-sampling
+# Apply SMOTE + ENN
 sm = SMOTEENN()
-X_resampled, y_resampled = sm.fit_transform(X, y)
+X_resampled, y_resampled = sm.fit_sample(X, y)
 X_res_vis = pca.transform(X_resampled)
 
 # Two subplots, unpack the axes array immediately

examples/combine/plot_smote_tomek.py

@@ -3,7 +3,7 @@
 SMOTE + Tomek
 =============
 
-An illustration of the random SMOTE + Tomek method.
+An illustration of the SMOTE + Tomek method.
 
 """
 
@@ -33,9 +33,9 @@
 # Fit and transform x to visualise inside a 2D feature space
 X_vis = pca.fit_transform(X)
 
-# Apply the random under-sampling
+# Apply SMOTE + Tomek links
 sm = SMOTETomek()
-X_resampled, y_resampled = sm.fit_transform(X, y)
+X_resampled, y_resampled = sm.fit_sample(X, y)
 X_res_vis = pca.transform(X_resampled)
 
 # Two subplots, unpack the axes array immediately

examples/ensemble/README.txt

@@ -1,6 +1,8 @@
 .. _ensemble_examples:
 
 Example using ensemble class methods
-===================================
+====================================
 
-Ensemble examples.
+Under-sampling methods implies that samples of the majority class are lost during the balancing procedure.
+Ensemble methods offer an alternative to use most of the samples.
+In fact, an ensemble of balanced sets is created and used to later train any classifier.

examples/ensemble/plot_balance_cascade.py

@@ -3,7 +3,7 @@
 Balance cascade
 ===============
 
-An illustration of the balance cascade method.
+An illustration of the balance cascade ensemble method.
 
 """
 
@@ -34,9 +34,9 @@
 # Fit and transform x to visualise inside a 2D feature space
 X_vis = pca.fit_transform(X)
 
-# Apply the random under-sampling
+# Apply Balance Cascade method
 bc = BalanceCascade()
-X_resampled, y_resampled = bc.fit_transform(X, y)
+X_resampled, y_resampled = bc.fit_sample(X, y)
 X_res_vis = []
 for X_res in X_resampled:
     X_res_vis.append(pca.transform(X_res))

examples/ensemble/plot_easy_ensemble.py

@@ -34,9 +34,9 @@
 # Fit and transform x to visualise inside a 2D feature space
 X_vis = pca.fit_transform(X)
 
-# Apply the random under-sampling
+# Apply Easy Ensemble
 ee = EasyEnsemble()
-X_resampled, y_resampled = ee.fit_transform(X, y)
+X_resampled, y_resampled = ee.fit_sample(X, y)
 X_res_vis = []
 for X_res in X_resampled:
     X_res_vis.append(pca.transform(X_res))

examples/over-sampling/README.txt

@@ -3,4 +3,4 @@
 Example using over-sampling class methods
 =========================================
 
-Over-sampling examples.
+Data balancing can be performed by over-sampling such that new samples are generated in the minority class to reach a given balancing ratio.

examples/over-sampling/plot_random_over_sampling.py

@@ -33,9 +33,9 @@
 # Fit and transform x to visualise inside a 2D feature space
 X_vis = pca.fit_transform(X)
 
-# Apply the random under-sampling
+# Apply the random over-sampling
 ros = RandomOverSampler()
-X_resampled, y_resampled = ros.fit_transform(X, y)
+X_resampled, y_resampled = ros.fit_sample(X, y)
 X_res_vis = pca.transform(X_resampled)
 
 # Two subplots, unpack the axes array immediately

examples/over-sampling/plot_smote.py

@@ -33,9 +33,9 @@
 # Fit and transform x to visualise inside a 2D feature space
 X_vis = pca.fit_transform(X)
 
-# Apply the random under-sampling
+# Apply regular SMOTE
 sm = SMOTE(kind='regular')
-X_resampled, y_resampled = sm.fit_transform(X, y)
+X_resampled, y_resampled = sm.fit_sample(X, y)
 X_res_vis = pca.transform(X_resampled)
 
 # Two subplots, unpack the axes array immediately