dowhy.gcm.ml 包#

子模块#

dowhy.gcm.ml.autogluon 模块#

dowhy.gcm.ml.classification 模块#

class dowhy.gcm.ml.classification.ClassificationModel[source]#

基类: PredictionModel

abstract property classes: List[str]#

abstract predict_probabilities(X: array) → ndarray[source]#

class dowhy.gcm.ml.classification.SklearnClassificationModel(sklearn_mdl: Any)[source]#

基类: SklearnRegressionModel, ClassificationModel

property classes: List[str]#

clone()[source]#: 使用相同的超参数克隆预测模型，但未拟合。:return: 预测模型的未拟合克隆。

predict_probabilities(X: array) → ndarray[source]#

class dowhy.gcm.ml.classification.SklearnClassificationModelWeighted(sklearn_mdl: Any)[source]#

基类: SklearnRegressionModelWeighted, ClassificationModel

property classes: List[str]#

clone()[source]#: 使用相同的超参数克隆预测模型，但未拟合。:return: 预测模型的未拟合克隆。

predict_probabilities(X: array) → ndarray[source]#

dowhy.gcm.ml.classification.create_ada_boost_classifier(**kwargs) → SklearnClassificationModel[source]#

dowhy.gcm.ml.classification.create_extra_trees_classifier(**kwargs) → SklearnClassificationModel[source]#

dowhy.gcm.ml.classification.create_gaussian_nb_classifier(**kwargs) → SklearnClassificationModel[source]#

dowhy.gcm.ml.classification.create_gaussian_process_classifier(**kwargs) → SklearnClassificationModel[source]#

dowhy.gcm.ml.classification.create_hist_gradient_boost_classifier(**kwargs) → SklearnClassificationModel[source]#

dowhy.gcm.ml.classification.create_knn_classifier(**kwargs) → SklearnClassificationModel[source]#

dowhy.gcm.ml.classification.create_logistic_regression_classifier(**kwargs) → SklearnClassificationModel[source]#

dowhy.gcm.ml.classification.create_polynom_logistic_regression_classifier(degree: int = 3, **kwargs_logistic_regression) → SklearnClassificationModel[source]#

dowhy.gcm.ml.classification.create_random_forest_classifier(**kwargs) → SklearnClassificationModel[source]#

dowhy.gcm.ml.classification.create_support_vector_classifier(**kwargs) → SklearnClassificationModel[source]#

dowhy.gcm.ml.prediction_model 模块#

class dowhy.gcm.ml.prediction_model.PredictionModel[source]#

基类: object

表示通用的预测模型实现。每个预测模型都应提供 fit 和 predict 方法。

abstract clone()[source]#

使用相同的超参数克隆预测模型，但未拟合。

返回:: 预测模型的未拟合克隆。

abstract fit(X: ndarray, Y: ndarray) → None[source]#

abstract predict(X: ndarray) → ndarray[source]#

dowhy.gcm.ml.regression 模块#

class dowhy.gcm.ml.regression.InvertibleExponentialFunction[source]#

基类: InvertibleFunction

evaluate(X: ndarray) → ndarray[source]#: 在输入上应用函数。

evaluate_inverse(X: ndarray) → ndarray[source]#: 返回在输入上应用函数逆的结果。

class dowhy.gcm.ml.regression.InvertibleFunction[source]#

基类: object

abstract evaluate(X: ndarray) → ndarray[source]#: 在输入上应用函数。

abstract evaluate_inverse(X: ndarray) → ndarray[source]#: 返回在输入上应用函数逆的结果。

class dowhy.gcm.ml.regression.InvertibleIdentityFunction[source]#

基类: InvertibleFunction

evaluate(X: ndarray) → ndarray[source]#: 在输入上应用函数。

evaluate_inverse(X: ndarray) → ndarray[source]#: 返回在输入上应用函数逆的结果。

class dowhy.gcm.ml.regression.InvertibleLogarithmicFunction[source]#

基类: InvertibleFunction

evaluate(X: ndarray) → ndarray[source]#: 在输入上应用函数。

evaluate_inverse(X: ndarray) → ndarray[source]#: 返回在输入上应用函数逆的结果。

class dowhy.gcm.ml.regression.LinearRegressionWithFixedParameter(coefficients: ndarray, intercept: float)[source]#

基类: PredictionModel

clone()[source]#

使用相同的超参数克隆预测模型，但未拟合。

返回:: 预测模型的未拟合克隆。

fit(X: ndarray, Y: ndarray) → None[source]#

predict(X: ndarray) → ndarray[source]#

class dowhy.gcm.ml.regression.SklearnRegressionModel(sklearn_mdl: Any)[source]#

基类: PredictionModel

sklearn 模型的通用包装类。

clone()[source]#: 使用相同的超参数克隆预测模型，但未拟合。:return: 预测模型的未拟合克隆。

fit(X: ndarray, Y: ndarray) → None[source]#

predict(X: array) → ndarray[source]#

property sklearn_model: Any#

class dowhy.gcm.ml.regression.SklearnRegressionModelWeighted(sklearn_mdl: Any)[source]#

基类: SklearnRegressionModel

fit(X: ndarray, Y: ndarray, sample_weight: ndarray | None = None) → None[source]#

dowhy.gcm.ml.regression.create_ada_boost_regressor(**kwargs) → SklearnRegressionModel[source]#

dowhy.gcm.ml.regression.create_elastic_net_regressor(**kwargs) → SklearnRegressionModel[source]#

dowhy.gcm.ml.regression.create_extra_trees_regressor(**kwargs) → SklearnRegressionModel[source]#

dowhy.gcm.ml.regression.create_gaussian_process_regressor(**kwargs) → SklearnRegressionModel[source]#

dowhy.gcm.ml.regression.create_hist_gradient_boost_regressor(**kwargs) → SklearnRegressionModel[source]#

dowhy.gcm.ml.regression.create_knn_regressor(**kwargs) → SklearnRegressionModel[source]#

dowhy.gcm.ml.regression.create_lasso_lars_ic_regressor(**kwargs) → SklearnRegressionModel[source]#

dowhy.gcm.ml.regression.create_lasso_regressor(**kwargs) → SklearnRegressionModel[source]#

dowhy.gcm.ml.regression.create_linear_regressor(**kwargs) → SklearnRegressionModel[source]#

dowhy.gcm.ml.regression.create_linear_regressor_with_given_parameters(coefficients: ndarray, intercept: float = 0) → LinearRegressionWithFixedParameter[source]#

dowhy.gcm.ml.regression.create_polynom_regressor(degree: int = 2, **kwargs_linear_model) → SklearnRegressionModel[source]#

dowhy.gcm.ml.regression.create_random_forest_regressor(**kwargs) → SklearnRegressionModel[source]#

dowhy.gcm.ml.regression.create_ridge_regressor(**kwargs) → SklearnRegressionModel[source]#

dowhy.gcm.ml.regression.create_support_vector_regressor(**kwargs) → SklearnRegressionModel[source]#

模块内容#

此模块定义了由不同的 FunctionalCausalModel 实现使用的 PredictionModel 实现，例如 PostNonlinearModel 或 AdditiveNoiseModel。