H2ONaiveBayesEstimator 产生一些错误
H2ONaiveBayesEstimator producing some errors
我正在尝试了解 H2ONaiveBayesEstimator 的工作原理。这是我的代码：
import pandas as pd
import h2o
from h2o.estimators import H2ONaiveBayesEstimator
df = pd.DataFrame(
{
"a": list(range(10)),
"b": list(range(10)),
"c": list(range(10))
}
)
data = df.values
h2o.init()
train_frame = h2o.H2OFrame(data)
nbm = H2ONaiveBayesEstimator()
y = train_frame.columns[-1]
nbm.train(y=y, training_frame=train_frame)
这是我得到的错误：
h2o.exceptions.H2OResponseError: ModelBuilderErrorV3 (water.exceptions.H2OModelBuilderIllegalArgumentException):
timestamp = 1568829213744
error_url = '/3/ModelBuilders/naivebayes'
msg = 'Illegal argument(s) for NaiveBayes model: NaiveBayes_model_python_1568829211691_1. Details: ERRR on field: _response: Response must be a categorical column'
dev_msg = 'Illegal argument(s) for NaiveBayes model: NaiveBayes_model_python_1568829211691_1. Details: ERRR on field: _response: Response must be a categorical column'
http_status = 412
values = {'messages': [{'_log_level': 5, '_field_name': '_keep_cross_validation_models', '_message': 'Only for cross-validation.'}, {'_log_level': 5, '_field_name': '_keep_cross_validation_predictions', '_message': 'Only for cross-validation.'}, {'_log_level': 5, '_field_name': '_keep_cross_validation_fold_assignment', '_message': 'Only for cross-validation.'}, {'_log_level': 5, '_field_name': '_fold_assignment', '_message': 'Only for cross-validation.'}, {'_log_level': 5, '_field_name': '_tweedie_power', '_message': 'Only for Tweedie Distribution.'}, {'_log_level': 5, '_field_name': '_tweedie_power', '_message': 'Tweedie power is only used for Tweedie distribution.'}, {'_log_level': 5, '_field_name': '_quantile_alpha', '_message': 'Quantile (alpha) is only used for Quantile regression.'}, {'_log_level': 5, '_field_name': '_max_after_balance_size', '_message': 'Balance classes is false, hide max_after_balance_size'}, {'_log_level': 5, '_field_name': '_balance_classes', '_message': 'Balance classes is only applicable to classification problems.'}, {'_log_level': 5, '_field_name': '_class_sampling_factors', '_message': 'Class sampling factors is only applicable to classification problems.'}, {'_log_level': 5, '_field_name': '_max_after_balance_size', '_message': 'Max after balance size is only applicable to classification problems.'}, {'_log_level': 5, '_field_name': '_max_confusion_matrix_size', '_message': 'Max confusion matrix size is only applicable to classification problems.'}, {'_log_level': 5, '_field_name': '_max_hit_ratio_k', '_message': 'Max K-value for hit ratio is only applicable to multi-class classification problems.'}, {'_log_level': 5, '_field_name': '_max_confusion_matrix_size', '_message': 'Only for multi-class classification problems.'}, {'_log_level': 5, '_field_name': '_max_after_balance_size', '_message': 'Only used with balanced classes'}, {'_log_level': 5, '_field_name': '_class_sampling_factors', '_message': 'Class sampling factors is only applicable if balancing classes.'}, {'_log_level': 1, '_field_name': '_response', '_message': 'Response must be a categorical column'}, {'_log_level': 5, '_field_name': '_balance_classes', '_message': 'Balance classes is not applicable to NaiveBayes.'}, {'_log_level': 5, '_field_name': '_class_sampling_factors', '_message': 'Class sampling factors is not applicable to NaiveBayes.'}, {'_log_level': 5, '_field_name': '_max_after_balance_size', '_message': 'Max after balance size is not applicable to NaiveBayes.'}], 'algo': 'NaiveBayes', 'parameters': {'_train': {'name': 'Key_Frame__upload_9cb9251216ede45dd909d3e9af81477a.hex', 'type': 'Key'}, '_valid': None, '_nfolds': 0, '_keep_cross_validation_models': True, '_keep_cross_validation_predictions': False, '_keep_cross_validation_fold_assignment': False, '_parallelize_cross_validation': True, '_auto_rebalance': True, '_seed': -1, '_fold_assignment': 'AUTO', '_categorical_encoding': 'AUTO', '_max_categorical_levels': 10, '_distribution': 'AUTO', '_tweedie_power': 1.5, '_quantile_alpha': 0.5, '_huber_alpha': 0.9, '_ignored_columns': None, '_ignore_const_cols': True, '_weights_column': None, '_offset_column': None, '_fold_column': None, '_check_constant_response': True, '_is_cv_model': False, '_score_each_iteration': False, '_max_runtime_secs': 0.0, '_stopping_rounds': 0, '_stopping_metric': 'AUTO', '_stopping_tolerance': 0.001, '_response_column': 'C3', '_balance_classes': False, '_max_after_balance_size': 5.0, '_class_sampling_factors': None, '_max_confusion_matrix_size': 20, '_checkpoint': None, '_pretrained_autoencoder': None, '_custom_metric_func': None, '_custom_distribution_func': None, '_export_checkpoints_dir': None, '_laplace': 0.0, '_eps_sdev': 0.0, '_min_sdev': 0.001, '_eps_prob': 0.0, '_min_prob': 0.001, '_compute_metrics': True}, 'error_count': 2}
exception_msg = 'Illegal argument(s) for NaiveBayes model: NaiveBayes_model_python_1568829211691_1. Details: ERRR on field: _response: Response must be a categorical column'
stacktrace =
.
.
.
parameters = {'__meta': {'schema_version': 3, 'schema_name': 'NaiveBayesParametersV3', 'schema_type': 'NaiveBayesParameters'}, 'model_id': None, 'training_frame': {'__meta': {'schema_version': 3, 'schema_name': 'FrameKeyV3', 'schema_type': 'Key<Frame>'}, 'name': 'Key_Frame__upload_9cb9251216ede45dd909d3e9af81477a.hex', 'type': 'Key<Frame>', 'URL': '/3/Frames/Key_Frame__upload_9cb9251216ede45dd909d3e9af81477a.hex'}, 'validation_frame': None, 'nfolds': 0, 'keep_cross_validation_models': True, 'keep_cross_validation_predictions': False, 'keep_cross_validation_fold_assignment': False, 'parallelize_cross_validation': True, 'distribution': 'AUTO', 'tweedie_power': 1.5, 'quantile_alpha': 0.5, 'huber_alpha': 0.9, 'response_column': {'__meta': {'schema_version': 3, 'schema_name': 'ColSpecifierV3', 'schema_type': 'VecSpecifier'}, 'column_name': 'C3', 'is_member_of_frames': None}, 'weights_column': None, 'offset_column': None, 'fold_column': None, 'fold_assignment': 'AUTO', 'categorical_encoding': 'AUTO', 'max_categorical_levels': 10, 'ignored_columns': None, 'ignore_const_cols': True, 'score_each_iteration': False, 'checkpoint': None, 'stopping_rounds': 0, 'max_runtime_secs': 0.0, 'stopping_metric': 'AUTO', 'stopping_tolerance': 0.001, 'custom_metric_func': None, 'custom_distribution_func': None, 'export_checkpoints_dir': None, 'balance_classes': False, 'class_sampling_factors': None, 'max_after_balance_size': 5.0, 'max_confusion_matrix_size': 20, 'max_hit_ratio_k': 0, 'laplace': 0.0, 'min_sdev': 0.001, 'eps_sdev': 0.0, 'min_prob': 0.001, 'eps_prob': 0.0, 'compute_metrics': True, 'seed': -1}
messages = [{'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'keep_cross_validation_models', 'message': 'Only for cross-validation.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'keep_cross_validation_predictions', 'message': 'Only for cross-validation.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'keep_cross_validation_fold_assignment', 'message': 'Only for cross-validation.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'fold_assignment', 'message': 'Only for cross-validation.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'tweedie_power', 'message': 'Only for Tweedie Distribution.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'tweedie_power', 'message': 'Tweedie power is only used for Tweedie distribution.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'quantile_alpha', 'message': 'Quantile (alpha) is only used for Quantile regression.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'max_after_balance_size', 'message': 'Balance classes is false, hide max_after_balance_size'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'balance_classes', 'message': 'Balance classes is only applicable to classification problems.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'class_sampling_factors', 'message': 'Class sampling factors is only applicable to classification problems.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'max_after_balance_size', 'message': 'Max after balance size is only applicable to classification problems.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'max_confusion_matrix_size', 'message': 'Max confusion matrix size is only applicable to classification problems.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'max_hit_ratio_k', 'message': 'Max K-value for hit ratio is only applicable to multi-class classification problems.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'max_confusion_matrix_size', 'message': 'Only for multi-class classification problems.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'max_after_balance_size', 'message': 'Only used with balanced classes'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'class_sampling_factors', 'message': 'Class sampling factors is only applicable if balancing classes.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'ERRR', 'field_name': 'response', 'message': 'Response must be a categorical column'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'balance_classes', 'message': 'Balance classes is not applicable to NaiveBayes.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'class_sampling_factors', 'message': 'Class sampling factors is not applicable to NaiveBayes.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'max_after_balance_size', 'message': 'Max after balance size is not applicable to NaiveBayes.'}]
error_count = 2
有人可以帮我弄清楚我的代码有什么问题吗？
朴素贝叶斯是一种分类（而非回归）算法，因此响应列必须是因子列。这对于 H2O 中的所有分类模型都是必需的。
您必须在训练模型之前执行此操作：
train_frame[y] = train_frame[y].asfactor()
我正在尝试了解 H2ONaiveBayesEstimator 的工作原理。这是我的代码：
import pandas as pd
import h2o
from h2o.estimators import H2ONaiveBayesEstimator
df = pd.DataFrame(
{
"a": list(range(10)),
"b": list(range(10)),
"c": list(range(10))
}
)
data = df.values
h2o.init()
train_frame = h2o.H2OFrame(data)
nbm = H2ONaiveBayesEstimator()
y = train_frame.columns[-1]
nbm.train(y=y, training_frame=train_frame)
这是我得到的错误：
h2o.exceptions.H2OResponseError: ModelBuilderErrorV3 (water.exceptions.H2OModelBuilderIllegalArgumentException):
timestamp = 1568829213744
error_url = '/3/ModelBuilders/naivebayes'
msg = 'Illegal argument(s) for NaiveBayes model: NaiveBayes_model_python_1568829211691_1. Details: ERRR on field: _response: Response must be a categorical column'
dev_msg = 'Illegal argument(s) for NaiveBayes model: NaiveBayes_model_python_1568829211691_1. Details: ERRR on field: _response: Response must be a categorical column'
http_status = 412
values = {'messages': [{'_log_level': 5, '_field_name': '_keep_cross_validation_models', '_message': 'Only for cross-validation.'}, {'_log_level': 5, '_field_name': '_keep_cross_validation_predictions', '_message': 'Only for cross-validation.'}, {'_log_level': 5, '_field_name': '_keep_cross_validation_fold_assignment', '_message': 'Only for cross-validation.'}, {'_log_level': 5, '_field_name': '_fold_assignment', '_message': 'Only for cross-validation.'}, {'_log_level': 5, '_field_name': '_tweedie_power', '_message': 'Only for Tweedie Distribution.'}, {'_log_level': 5, '_field_name': '_tweedie_power', '_message': 'Tweedie power is only used for Tweedie distribution.'}, {'_log_level': 5, '_field_name': '_quantile_alpha', '_message': 'Quantile (alpha) is only used for Quantile regression.'}, {'_log_level': 5, '_field_name': '_max_after_balance_size', '_message': 'Balance classes is false, hide max_after_balance_size'}, {'_log_level': 5, '_field_name': '_balance_classes', '_message': 'Balance classes is only applicable to classification problems.'}, {'_log_level': 5, '_field_name': '_class_sampling_factors', '_message': 'Class sampling factors is only applicable to classification problems.'}, {'_log_level': 5, '_field_name': '_max_after_balance_size', '_message': 'Max after balance size is only applicable to classification problems.'}, {'_log_level': 5, '_field_name': '_max_confusion_matrix_size', '_message': 'Max confusion matrix size is only applicable to classification problems.'}, {'_log_level': 5, '_field_name': '_max_hit_ratio_k', '_message': 'Max K-value for hit ratio is only applicable to multi-class classification problems.'}, {'_log_level': 5, '_field_name': '_max_confusion_matrix_size', '_message': 'Only for multi-class classification problems.'}, {'_log_level': 5, '_field_name': '_max_after_balance_size', '_message': 'Only used with balanced classes'}, {'_log_level': 5, '_field_name': '_class_sampling_factors', '_message': 'Class sampling factors is only applicable if balancing classes.'}, {'_log_level': 1, '_field_name': '_response', '_message': 'Response must be a categorical column'}, {'_log_level': 5, '_field_name': '_balance_classes', '_message': 'Balance classes is not applicable to NaiveBayes.'}, {'_log_level': 5, '_field_name': '_class_sampling_factors', '_message': 'Class sampling factors is not applicable to NaiveBayes.'}, {'_log_level': 5, '_field_name': '_max_after_balance_size', '_message': 'Max after balance size is not applicable to NaiveBayes.'}], 'algo': 'NaiveBayes', 'parameters': {'_train': {'name': 'Key_Frame__upload_9cb9251216ede45dd909d3e9af81477a.hex', 'type': 'Key'}, '_valid': None, '_nfolds': 0, '_keep_cross_validation_models': True, '_keep_cross_validation_predictions': False, '_keep_cross_validation_fold_assignment': False, '_parallelize_cross_validation': True, '_auto_rebalance': True, '_seed': -1, '_fold_assignment': 'AUTO', '_categorical_encoding': 'AUTO', '_max_categorical_levels': 10, '_distribution': 'AUTO', '_tweedie_power': 1.5, '_quantile_alpha': 0.5, '_huber_alpha': 0.9, '_ignored_columns': None, '_ignore_const_cols': True, '_weights_column': None, '_offset_column': None, '_fold_column': None, '_check_constant_response': True, '_is_cv_model': False, '_score_each_iteration': False, '_max_runtime_secs': 0.0, '_stopping_rounds': 0, '_stopping_metric': 'AUTO', '_stopping_tolerance': 0.001, '_response_column': 'C3', '_balance_classes': False, '_max_after_balance_size': 5.0, '_class_sampling_factors': None, '_max_confusion_matrix_size': 20, '_checkpoint': None, '_pretrained_autoencoder': None, '_custom_metric_func': None, '_custom_distribution_func': None, '_export_checkpoints_dir': None, '_laplace': 0.0, '_eps_sdev': 0.0, '_min_sdev': 0.001, '_eps_prob': 0.0, '_min_prob': 0.001, '_compute_metrics': True}, 'error_count': 2}
exception_msg = 'Illegal argument(s) for NaiveBayes model: NaiveBayes_model_python_1568829211691_1. Details: ERRR on field: _response: Response must be a categorical column'
stacktrace =
.
.
.
parameters = {'__meta': {'schema_version': 3, 'schema_name': 'NaiveBayesParametersV3', 'schema_type': 'NaiveBayesParameters'}, 'model_id': None, 'training_frame': {'__meta': {'schema_version': 3, 'schema_name': 'FrameKeyV3', 'schema_type': 'Key<Frame>'}, 'name': 'Key_Frame__upload_9cb9251216ede45dd909d3e9af81477a.hex', 'type': 'Key<Frame>', 'URL': '/3/Frames/Key_Frame__upload_9cb9251216ede45dd909d3e9af81477a.hex'}, 'validation_frame': None, 'nfolds': 0, 'keep_cross_validation_models': True, 'keep_cross_validation_predictions': False, 'keep_cross_validation_fold_assignment': False, 'parallelize_cross_validation': True, 'distribution': 'AUTO', 'tweedie_power': 1.5, 'quantile_alpha': 0.5, 'huber_alpha': 0.9, 'response_column': {'__meta': {'schema_version': 3, 'schema_name': 'ColSpecifierV3', 'schema_type': 'VecSpecifier'}, 'column_name': 'C3', 'is_member_of_frames': None}, 'weights_column': None, 'offset_column': None, 'fold_column': None, 'fold_assignment': 'AUTO', 'categorical_encoding': 'AUTO', 'max_categorical_levels': 10, 'ignored_columns': None, 'ignore_const_cols': True, 'score_each_iteration': False, 'checkpoint': None, 'stopping_rounds': 0, 'max_runtime_secs': 0.0, 'stopping_metric': 'AUTO', 'stopping_tolerance': 0.001, 'custom_metric_func': None, 'custom_distribution_func': None, 'export_checkpoints_dir': None, 'balance_classes': False, 'class_sampling_factors': None, 'max_after_balance_size': 5.0, 'max_confusion_matrix_size': 20, 'max_hit_ratio_k': 0, 'laplace': 0.0, 'min_sdev': 0.001, 'eps_sdev': 0.0, 'min_prob': 0.001, 'eps_prob': 0.0, 'compute_metrics': True, 'seed': -1}
messages = [{'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'keep_cross_validation_models', 'message': 'Only for cross-validation.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'keep_cross_validation_predictions', 'message': 'Only for cross-validation.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'keep_cross_validation_fold_assignment', 'message': 'Only for cross-validation.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'fold_assignment', 'message': 'Only for cross-validation.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'tweedie_power', 'message': 'Only for Tweedie Distribution.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'tweedie_power', 'message': 'Tweedie power is only used for Tweedie distribution.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'quantile_alpha', 'message': 'Quantile (alpha) is only used for Quantile regression.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'max_after_balance_size', 'message': 'Balance classes is false, hide max_after_balance_size'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'balance_classes', 'message': 'Balance classes is only applicable to classification problems.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'class_sampling_factors', 'message': 'Class sampling factors is only applicable to classification problems.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'max_after_balance_size', 'message': 'Max after balance size is only applicable to classification problems.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'max_confusion_matrix_size', 'message': 'Max confusion matrix size is only applicable to classification problems.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'max_hit_ratio_k', 'message': 'Max K-value for hit ratio is only applicable to multi-class classification problems.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'max_confusion_matrix_size', 'message': 'Only for multi-class classification problems.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'max_after_balance_size', 'message': 'Only used with balanced classes'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'class_sampling_factors', 'message': 'Class sampling factors is only applicable if balancing classes.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'ERRR', 'field_name': 'response', 'message': 'Response must be a categorical column'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'balance_classes', 'message': 'Balance classes is not applicable to NaiveBayes.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'class_sampling_factors', 'message': 'Class sampling factors is not applicable to NaiveBayes.'}, {'__meta': {'schema_version': 3, 'schema_name': 'ValidationMessageV3', 'schema_type': 'ValidationMessage'}, 'message_type': 'TRACE', 'field_name': 'max_after_balance_size', 'message': 'Max after balance size is not applicable to NaiveBayes.'}]
error_count = 2
有人可以帮我弄清楚我的代码有什么问题吗？
朴素贝叶斯是一种分类（而非回归）算法，因此响应列必须是因子列。这对于 H2O 中的所有分类模型都是必需的。
您必须在训练模型之前执行此操作：
train_frame[y] = train_frame[y].asfactor()