GridSearchCV()：ValueError：输入包含 NaN、无穷大或对于 dtype('float64') 来说太大的值

Posted 2023-03-12

【中文标题】GridSearchCV()：ValueError：输入包含 NaN、无穷大或对于 dtype(\'float64\') 来说太大的值

【英文标题】：GridSearchCV(): ValueError: Input contains NaN, infinity or a value too large for dtype('float64')GridSearchCV()：ValueError：输入包含 NaN、无穷大或对于 dtype('float64') 来说太大的值

【发布时间】：2020-08-16 00:51:39

【问题描述】：

当我尝试在 MLP 分类器上执行 GridsearchCV 时，标题中出现 ValueError。当然，我检查了我的数据集中是否存在任何 np.inf 或 np.nan，但它们不存在：

    print(np.any(np.isnan(X)))

返回假

    print(np.all(np.isfinite(X)))

返回真

我还将所有值都转换为 np.float64

X = X.values.astype(np.float64)
Y = Y.values

我的 scikit-learn 版本是 0.22.2.post1（最新）

我正在尝试执行的代码：

from scipy.stats import randint as sp_randint

hiddenlayers = [(sp_randint.rvs(100,600,1),sp_randint.rvs(100,600,1),), (sp_randint.rvs(100,600,1),)]
alpha_range = 10.0 ** np.arange(-2, 1)


param_grid_MLP = ['solver': ['lbfgs'],
                   'hidden_layer_sizes': hiddenlayers,
                   'activation': ['identity','tanh', 'relu', 'logistic'],
                   'alpha': alpha_range
                  ,
                 'solver': ['sgd'],
                  'hidden_layer_sizes': hiddenlayers,
                   'activation': ['identity','tanh', 'relu', 'logistic'],
                   'alpha': alpha_range,
                  'learning_rate':['constant','invscaling','adaptive']
                  ,
                 'solver': ['adam'],
                  'hidden_layer_sizes': hiddenlayers,
                   'activation': ['identity','tanh', 'relu', 'logistic'],
                   'alpha': alpha_range
                  ]

mlp = MLPClassifier(random_state=0)
cross_validation = StratifiedKFold(5)

# scoring = 'AUC': 'roc_auc', 
#            'Accuracy': make_scorer(accuracy_score),
#            'Recall':make_scorer(recall_score,pos_label='crafted'),
#            'Precision': make_scorer(precision_score,pos_label='crafted')

scoring = 'AUC': 'roc_auc', 
           'Accuracy': make_scorer(accuracy_score),
            'Recall':make_scorer(recall_score,pos_label='crafted')

grid_search_MLP = GridSearchCV(estimator=mlp, 
            param_grid=param_grid_MLP, 
            scoring=scoring,cv=cross_validation.split(X_train,y_train),
            refit='Recall',
            n_jobs=-1,
            verbose=True)

grid_search_MLP.fit(X_train,y_train)

print('Best score: '.format(grid_search_MLP.best_score_))
print('Best index: '.format(grid_search_MLP.best_index_))
print('Best parameters: '.format(grid_search_MLP.best_params_))

mlp = grid_search_MLP.best_estimator_
mlp

完整的错误回溯：

Traceback (most recent call last):
  File "/usr/local/lib/python3.7/dist-packages/joblib/externals/loky/process_executor.py", line 418, in _process_worker
    r = call_item()
  File "/usr/local/lib/python3.7/dist-packages/joblib/externals/loky/process_executor.py", line 272, in __call__
    return self.fn(*self.args, **self.kwargs)
  File "/usr/local/lib/python3.7/dist-packages/joblib/_parallel_backends.py", line 608, in __call__
    return self.func(*args, **kwargs)
  File "/usr/local/lib/python3.7/dist-packages/joblib/parallel.py", line 256, in __call__
    for func, args, kwargs in self.items]
  File "/usr/local/lib/python3.7/dist-packages/joblib/parallel.py", line 256, in <listcomp>
    for func, args, kwargs in self.items]
  File "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_validation.py", line 544, in _fit_and_score
    test_scores = _score(estimator, X_test, y_test, scorer)
  File "/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_validation.py", line 591, in _score
    scores = scorer(estimator, X_test, y_test)
  File "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_scorer.py", line 87, in __call__
    *args, **kwargs)
  File "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_scorer.py", line 332, in _score
    return self._sign * self._score_func(y, y_pred, **self._kwargs)
  File "/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_ranking.py", line 369, in roc_auc_score
    y_score = check_array(y_score, ensure_2d=False)
  File "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py", line 578, in check_array
    allow_nan=force_all_finite == 'allow-nan')
  File "/usr/local/lib/python3.7/dist-packages/sklearn/utils/validation.py", line 60, in _assert_all_finite
    msg_dtype if msg_dtype is not None else X.dtype)
ValueError: Input contains NaN, infinity or a value too large for dtype('float64').

【问题讨论】：

你从哪里得到你的阵列？一个csv文件？你如何建造它？你能打印出来吗？

【参考方案1】：

在我看来，您的数组中的值可能已损坏，或者是非数值。在转换为浮点数之前，尝试检查数组中是否还有其他类型。还尝试找到数组中的最小值和最大值，这可能有助于找到引发错误的值。

【参考方案2】：

尝试给出一个大数字，或者一个一个地运行网格的 3 个部分。如果你意识到sgd 给出了问题，它可能在这里解释MLPRegressor error when solver sgd is used