ITMO_FS.wrappers.deterministic.AddDelWrapper

class ITMO_FS.wrappers.deterministic.AddDelWrapper(estimator, score, maximize=True, seed=42)

Creates add-del feature wrapper

Parameters:

• estimator (object) – A supervised learning estimator with a fit method
• score (boolean) – A callable function which will be used to estimate score
• score – maximize = True if bigger values are better for score function
• seed (int) – Seed for python random
• best_score (float) – The best score of given metric on the feature combination after add-del procedure

See also

Lecture, p.13

Examples

>>> from sklearn.metrics import accuracy_score
>>> from sklearn import datasets,linear_model
>>> data = datasets.make_classification(n_samples=1000, n_features=20)
>>> X = np.array(data[0])
>>> y = np.array(data[1])
>>> lg = linear_model.LogisticRegression(solver='lbfgs')
>>> add_del = AddDelWrapper(lg, accuracy_score)
>>> add_del.fit(X, y)

>>> from sklearn.metrics import mean_absolute_error
>>> boston = datasets.load_boston()
>>> X = boston['data']
>>> y = boston['target']
>>> lasso = linear_model.Lasso()
>>> add_del = AddDelWrapper(lasso, mean_absolute_error, maximize=False)
>>> add_del.fit(X, y)

__init__(estimator, score, maximize=True, seed=42)

Initialize self. See help(type(self)) for accurate signature.

fit(X, y, cv=3, silent=True)

Fits wrapper.

Parameters:

• X (numpy array or pandas DataFrame, shape (n_samples, n_features)) – The training input samples.
• y (numpy array of pandas Series, shape (n_samples, )) – The target values.
• cv=3 (int) – Number of splits in cross-validation
• silent=True (boolean) – If silent=False then prints all the scores during add-del procedure
• Returns –
• ---------- –
• features (list) – List of feature after add-del procedure

Examples

Parameters:

• silent –
• y –
• X –
• cv –