abalone.py 文件源码

def load_abalone_data(proportion=1044./4177):
    from sklearn import datasets
    from sklearn import preprocessing
    from sklearn import cross_validation
    abalone = datasets.fetch_mldata('regression-datasets abalone')
    X_cate = np.array([abalone.target[i].tolist()
        for i in range(abalone.target.shape[0])])
    X_cate = preprocessing.label_binarize(X_cate, np.unique(X_cate))
    X = np.hstack((X_cate, abalone.data))
    y = abalone.int1[0].T.astype(np.float64)
    y = y[:, None]
    X = X.astype(np.float64)
    X_train, X_test, y_train, y_test = \
        cross_validation.train_test_split(X, y, test_size=proportion)
    return X_train, y_train, X_test, y_test