Contact
CoCalc Logo Icon
StoreFeaturesDocsShareSupport News AboutSign UpSign In
| Download
Project: Testing 18.04
Views: 407
Kernel: Python 3 (system-wide)

imblearn on cocalc

kernel: python 3 ubuntu linux

import imblearn imblearn.__version__
'0.7.0'
from collections import Counter from sklearn.datasets import make_classification from imblearn.over_sampling import SMOTE # doctest: +NORMALIZE_WHITESPACE
# implementation of SMOTE - Synthetic Minority Over-sampling Technique
X, y = make_classification( n_classes=2, class_sep=2, weights=[0.1, 0.9], n_informative=3, n_redundant=1, flip_y=0, n_features=20, n_clusters_per_class=1, n_samples=1000, random_state=10) print('Original dataset shape %s' % Counter(y))
Original dataset shape Counter({1: 900, 0: 100})
sm = SMOTE(random_state=42) sm
SMOTE(random_state=42)
X_res, y_res = sm.fit_resample(X, y) print('Resampled dataset shape %s' % Counter(y_res))
Resampled dataset shape Counter({0: 900, 1: 900})