def __init__(self,
pool_classifiers=None,
k=7,
DFP=False,
with_IH=False,
safe_k=None,
IH_rate=0.30,
needs_proba=False,
random_state=None,
knn_classifier='knn',
DSEL_perc=0.5,
knne=False):
self.pool_classifiers = pool_classifiers
self.k = k
self.DFP = DFP
self.with_IH = with_IH
self.safe_k = safe_k
self.IH_rate = IH_rate
self.needs_proba = needs_proba
self.random_state = random_state
self.knn_classifier = knn_classifier
self.DSEL_perc = DSEL_perc
self.knne = knne
# Check optional dependency
if knn_classifier == 'faiss' and not faiss_knn_wrapper.is_available():
raise ImportError(
'Using knn_classifier="faiss" requires that the FAISS library '
'be installed.Please check the Installation Guide.')
def _set_knn_type(self):
if self.knn_classifier == 'faiss':
if not faiss_knn_wrapper.is_available():
raise ImportError(
'Using knn_classifier="faiss" requires that the FAISS '
'library be installed.Please check the Installation '
'Guide.')
self.knn_type_ = faiss_knn_wrapper.FaissKNNClassifier
else:
self.knn_type_ = KNeighborsClassifier
def _set_knn_type(self):
if self.knn_classifier is None or self.knn_classifier in ['knn',
'sklearn']:
self.knn_type_ = KNeighborsClassifier
elif self.knn_classifier == 'faiss':
if not faiss_knn_wrapper.is_available():
raise ImportError(
'Using knn_classifier="faiss" requires that the FAISS '
'library be installed.Please check the Installation '
'Guide.')
self.knn_type_ = faiss_knn_wrapper.FaissKNNClassifier
elif callable(self.knn_classifier):
self.knn_type_ = self.knn_classifier
else:
raise ValueError('"knn_classifier" should be one of the following '
'["knn", "faiss", None] or an estimator class.')
pool_classifiers, X_dsel, y_dsel, X_test, y_test = setup_classifiers()
kne = KNORAE(pool_classifiers)
params = {'k': [1, 3, 5, 7]}
grid = GridSearchCV(kne, params)
grid.fit(X_dsel, y_dsel)
grid.best_estimator_.score(X_test, y_test)
knn_methods = [None]
if faiss_knn_wrapper.is_available():
knn_methods.append(faiss_knn_wrapper.FaissKNNClassifier)
else:
warnings.warn("Not testing FAISS for KNN")
def test_label_encoder_integration_list_classifiers():
rng = np.random.RandomState(123456)
X_dsel, X_test, X_train, y_dsel, y_test, y_train = load_dataset(
encode_labels=['no', 'yes'], rng=rng)
pool_classifiers = [LogisticRegression(), SVC(probability=True)]
[clf.fit(X_train, y_train) for clf in pool_classifiers]
knorau = KNORAU(pool_classifiers)
knorau.fit(X_dsel, y_dsel)
import pytest
from sklearn.neighbors import KNeighborsClassifier
from deslib.tests.examples_test import *
from deslib.tests.test_des_integration import load_dataset
from deslib.util import faiss_knn_wrapper
@pytest.mark.skipif(not faiss_knn_wrapper.is_available(),
reason="requires the faiss library")
def test_faiss_predict():
rng = np.random.RandomState(123456)
_, X_test, X_train, _, _, y_train = load_dataset(None, rng)
k = 7
X_train = X_train.astype(np.float32)
X_test = X_test.astype(np.float32)
f_knn_test = faiss_knn_wrapper.FaissKNNClassifier(n_neighbors=k)
f_knn_test.fit(X_train, y_train)
f_knn_preds = f_knn_test.predict(X_test)
knn_test = KNeighborsClassifier(n_neighbors=k)
knn_test.fit(X_train, y_train)
knn_preds = knn_test.predict(X_test)
assert ((f_knn_preds - knn_preds).sum() == 0)
