DatasetProfileRunConfig 类

注意

这是一个实验性的类,可能会在任何时候更改。 有关详细信息,请参阅 https://aka.ms/azuremlexperimental

表示用于提交 Azure 机器学习中的数据集配置文件运行的配置。

此配置对象包含并保留用于配置试验运行的参数,以及运行时使用的数据。

继承
builtins.object
DatasetProfileRunConfig

构造函数

DatasetProfileRunConfig(dataset, compute_target='local', datastore_name=None)

参数

dataset
TabularDataset
必需

作为此配置文件运行目标的表格数据集。 有关表格数据集的详细信息,请参阅 https://docs.microsoft.com/en-us/python/api/azureml-core/azureml.data.tabulardataset

compute_target
Union[ComputeTarget, str]
必需

用于运行自动化机器学习试验的 Azure 机器学习计算目标。 指定“local”以使用本地计算。 默认值为“local”。 有关计算目标的详细信息,请参阅 https://docs.microsoft.com/en-us/python/api/azureml-core/azureml.core.computetarget

注解

下面的代码演示了一个基本示例,它创建 DatasetProfileRunConfig 对象并在“local”上提交配置文件计算试验:


   workspace = Workspace.from_config()
   dataset = Dataset.get_by_name(workspace, name='test_dataset')
   dprc = DatasetProfileRunConfig(dataset=dataset, compute_target='local')
   exp = Experiment(workspace, "profile_experiment_run")
   profile_run = exp.submit(dprc)
   profile_run.run.wait_for_completion(raise_on_error=True, wait_post_processing=True)
   profile = profile_run.get_profile()