def test_sample_kmeans_not_enough_clusters():
rng = np.random.RandomState(42)
X = rng.randn(30, 2)
y = np.array([1] * 20 + [0] * 10)
smote = KMeansSMOTE(random_state=42, kmeans_estimator=30, k_neighbors=2)
with pytest.raises(RuntimeError):
smote.fit_resample(X, y)
def test_kmeans_smote_param_error(data, density_exponent, cluster_balance_threshold):
X, y = data
kmeans_smote = KMeansSMOTE(
density_exponent=density_exponent,
cluster_balance_threshold=cluster_balance_threshold,
)
with pytest.raises(ValueError, match="should be 'auto' when a string"):
kmeans_smote.fit_resample(X, y)
def test_sample_kmeans_density_estimation(data, density_exponent,
cluster_balance_threshold):
X, y = data
smote = KMeansSMOTE(
random_state=42,
density_exponent=density_exponent,
cluster_balance_threshold=cluster_balance_threshold,
)
smote.fit_resample(X, y)
def test_sample_kmeans_density_estimation(density_exponent, cluster_balance_threshold):
X, y = make_classification(
n_samples=10_000, n_classes=2, weights=[0.3, 0.7], random_state=42
)
smote = KMeansSMOTE(
random_state=0,
density_exponent=density_exponent,
cluster_balance_threshold=cluster_balance_threshold,
)
smote.fit_resample(X, y)
def sample(xtrain, ytrain):
sm = KMeansSMOTE(random_state=42)
x_res, y_res = sm.fit_resample(xtrain, ytrain)
y = y_res
x_res = pd.DataFrame(x_res)
#y_res=pd.DataFrame(y_res)
x_res.columns = xtrain.columns
#y_res.columns=["Leak_type"]
return x_res, y_res
def over_sample_data(matrix, y_train):
add_to_log('Over Sampling')
add_to_log('Sample distribution %s' % Counter(y_train))
b_line = KMeansSMOTE(k_neighbors=5,
sampling_strategy='not majority',
n_jobs=-1,
random_state=3,
kmeans_estimator=100)
matrix_resampled, y_resampled = b_line.fit_resample(matrix, y_train)
add_to_log('Resample distribution %s' % Counter(y_resampled))
return matrix_resampled, y_resampled
def test_sample_kmeans_custom(data, k_neighbors, kmeans_estimator):
X, y = data
kmeans_smote = KMeansSMOTE(
random_state=42,
kmeans_estimator=kmeans_estimator,
k_neighbors=k_neighbors,
)
X_resampled, y_resampled = kmeans_smote.fit_resample(X, y)
assert X_resampled.shape == (24, 2)
assert y_resampled.shape == (24,)
assert kmeans_smote.nn_k_.n_neighbors == 3
assert kmeans_smote.kmeans_estimator_.n_clusters == 3
def keans_smote(X,
y,
visualize=False,
pca2d=True,
pca3d=True,
tsne=True,
pie_evr=True):
sm = KMeansSMOTE(random_state=42)
X_res, y_res = sm.fit_resample(X, y)
if visualize == True:
hist_over_and_undersampling(y_res)
pca_general(X_res, y_res, d2=pca2d, d3=pca3d, pie_evr=pie_evr)
return X_res, y_res
def test_kmeans_smote(data):
X, y = data
kmeans_smote = KMeansSMOTE(
kmeans_estimator=1,
random_state=42,
cluster_balance_threshold=0.0,
k_neighbors=5,
)
smote = SMOTE(random_state=42)
X_res_1, y_res_1 = kmeans_smote.fit_resample(X, y)
X_res_2, y_res_2 = smote.fit_resample(X, y)
assert_allclose(X_res_1, X_res_2)
assert_array_equal(y_res_1, y_res_2)
assert kmeans_smote.nn_k_.n_neighbors == 6
assert kmeans_smote.kmeans_estimator_.n_clusters == 1
assert "batch_size" in kmeans_smote.kmeans_estimator_.get_params()
def test_sample_kmeans_not_enough_clusters(data):
X, y = data
smote = KMeansSMOTE(cluster_balance_threshold=10, random_state=42)
with pytest.raises(RuntimeError):
smote.fit_resample(X, y)
#
# border_sm = BorderlineSMOTE(k_neighbors=27, random_state=91, sampling_strategy=1)
#
# sm = SVMSMOTE(random_state=91, k_neighbors=2, sampling_strategy=1, svm_estimator=SVM_smote)
#
# ada = ADASYN(random_state=91, n_neighbors=27, sampling_strategy=1, n_jobs=6)
Kmeans = KMeansSMOTE(random_state=91,
k_neighbors=2,
sampling_strategy=1,
n_jobs=6,
kmeans_estimator=MiniBatchKMeans(n_clusters=20))
'''Muestreo Sintetico'''
# Xtrain, ytrain = SMOTE().fit_resample(Xtrain, ytrain)
Xtrain, ytrain = Kmeans.fit_resample(Xtrain, ytrain)
X_train, X_val, y_train, y_val = train_test_split(Xtrain,
ytrain,
test_size=0.33,
random_state=42)
'''Selección de caracteristicas'''
# rel_MI = SelectKBest(score_func=score_func, k=num_features)
# Xtrain = rel_MI.fit_transform(Xtrain, ytrain)
# Xtest = rel_MI.transform(Xtest)
# rel_MI_support = rel_MI.get_support()
# rel_MI_feature = X_frame.loc[:, rel_MI_support].columns.tolist()
# rel_MI_scores = rel_MI.scores_[rel_MI_support].tolist()
# feature_selection_df = pd.DataFrame({'Feature': rel_MI_feature, 'Score':rel_MI_scores})
def keans_smote(X, y):
sm = KMeansSMOTE(random_state=42)
X_res, y_res = sm.fit_resample(X, y)
return X_res, y_res