Azure ML 表格数据集:缺少 1 个必需的位置参数:'stream_column'

Azure ML Tabular Dataset : missing 1 required positional argument: 'stream_column'

对于Python API for tabular dataset of AzureML (azureml.data.TabularDataset),已经介绍了两种实验方法:

  1. download(stream_column, target_path=None, overwrite=False, ignore_not_found=True)
  2. mount(stream_column, mount_point=None)

参数 stream_column 已定义为要装载或下载的流列。

stream_column的实际含义是什么?我在哪里看不到任何例子?

任何指针都会有所帮助。

堆栈跟踪:

Method download: This is an experimental method, and may change at any time. Please see https://aka.ms/azuremlexperimental for more information.
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
/tmp/ipykernel_11561/3904436543.py in <module>
----> 1 tab_dataset.download(target_path="../data/tabular")

/anaconda/envs/azureml_py38/lib/python3.8/site-packages/azureml/_base_sdk_common/_docstring_wrapper.py in wrapped(*args, **kwargs)
     50     def wrapped(*args, **kwargs):
     51         module_logger.warning("Method {0}: {1} {2}".format(func.__name__, _method_msg, _experimental_link_msg))
---> 52         return func(*args, **kwargs)
     53     return wrapped
     54 

/anaconda/envs/azureml_py38/lib/python3.8/site-packages/azureml/data/_loggerfactory.py in wrapper(*args, **kwargs)
    130             with _LoggerFactory.track_activity(logger, func.__name__, activity_type, custom_dimensions) as al:
    131                 try:
--> 132                     return func(*args, **kwargs)
    133                 except Exception as e:
    134                     if hasattr(al, 'activity_info') and hasattr(e, 'error_code'):

TypeError: download() missing 1 required positional argument: 'stream_column'

2022 年 3 月 5 日更新

我将此作为 Azure 的支持票发布。以下是我收到的答复:

As you can see from our documentation of TabularDataset Class, the “stream_column” parameter is required. So, that error is occurring because you are not passing any parameters when you are calling the download method. The “stream_column” parameter should have the stream column to download/mount. So, you need to pass the column name that contains the paths from which the data will be streamed.
Please find an example here.