问题描述
我是包操作的新手。
我找到了这个例子:
我测试了一下。我想将每个 automl 试验的输出保存在 Trial 对象 中。我还想获取和设置每个 Trial 的超参数。为了设置 automl 对象的超参数,我使用了以下 python 代码:auto_ml.get_hyperparams()['Pipeline']
这是输出:
HyperparameterSamples([('choice','SKLearnWrapper_DecisionTreeClassifier'),('SKLearnWrapper_DecisionTreeClassifier',HyperparameterSamples([('enabled',True),('Optional(SKLearnWrapper_DecisionTreeClassifier)',HyperparameterSamples(['),('class_weight',None),('criterion','gini'),('max_depth',('max_features',('max_leaf_nodes',('min_impurity_decrease',0.0),('min_impurity_split',('min_samples_leaf',1),('min_samples_split',2),('min_weight_fraction_leaf',('random_state',('splitter','best') ]))])),('SKLearnWrapper_ExtraTreeClassifier',False),('Optional(SKLearnWrapper_ExtraTreeClassifier)',HyperparameterSamples([('ccp_alpha','auto'),None) ),('min_weight_frac tion_leaf','random')]))]),('RidgeClassifier',('Optional( RidgeClassifier)',HyperparameterSamples([('OutputTransformerWrapper',HyperparameterSamples([('NumpyRavel',HyperparameterSamples())])),('SKLearnWrapper_RidgeClassifier',HyperparameterSamples([('alpha',1.0),('copy_X',('fit_intercept',('max_iter',('normalize',('solver','auto' ),('tol',0.001)]))])]),('LogisticRegression',('Optional(LogisticRegression)',HyperparameterSamples([('OutputTransformerWrapper'),('SKLearnWrapper_LogisticRegression',HyperparameterSamples([('C',('dual',( 'fit_intercept',('intercept_scaling',('l1_ratio',100),('multi_class',('n_jobs',('penalty','l2'),'lbfgs'),0.0001),('verbose',0),('warm_start',False )]))]))])),('RandomForestClassifier',('Optional(RandomForestClassifier)',HyperparameterSamples([('NumpyRavel'),HyperparameterSamples())]),('SKLearnWrapper_RandomForestClassifier',HyperparameterSamples([('bootstrap',('ccp_alpha',('max_samples',('n_estimators',('oob_score',( 'random_state',False)]))]))]),('joiner',HyperparameterSamples())])
输出是 HyperparameterSamples 对象,我想把它转换成 Trials,这可能吗?
解决方法
暂无找到可以解决该程序问题的有效方法,小编努力寻找整理中!
如果你已经找到好的解决方法,欢迎将解决方案带上本链接一起发送给小编。
小编邮箱:dio#foxmail.com (将#修改为@)