TypeError: argument of type 'azureml.dataprep.rslex.StreamInfo' is not iterable

まず、提供されている NycTlcYellow の再現手順を azureml-opendatasets==1.55.0 で実行し、次に azureml/opendatasets/dataaccess/_blob_accessor.py、特に get_spark_dataframe と _get_relative_path を調査します。target_paths に StreamInfo が含まれる理由を追跡し、pandas パスと Spark パスを比較します。完了の条件は、報告された TypeError が発生せずに to_spark_dataframe が完了することです。

Hi, I encountered below error, I have azureml-opendatasets==1.55.0

Calling to_spark_dataframe()
Traceback (most recent call last):
File "/mnt/c/users/cruiseli/OneDrive - Microsoft/Desktop/workspace/SynapseML-Utils/test_aml.py", line 30, in
nyc_tlc_df2 = nyc_tlc.to_spark_dataframe()
File "/home/cruise/mambaforge/lib/python3.10/site-packages/azureml/opendatasets/accessories/_loggerfactory.py", line 139, in wrapper
return func(*args, **kwargs)
File "/home/cruise/mambaforge/lib/python3.10/site-packages/azureml/opendatasets/accessories/open_dataset_base.py", line 164, in to_spark_dataframe
return self._to_spark_dataframe()
File "/home/cruise/mambaforge/lib/python3.10/site-packages/azureml/opendatasets/accessories/open_dataset_base.py", line 305, in _to_spark_dataframe
return self._blob_accessor.get_spark_dataframe(
File "/home/cruise/mambaforge/lib/python3.10/site-packages/azureml/opendatasets/dataaccess/_blob_accessor.py", line 303, in get_spark_dataframe
paths = [wasab_format % (self._blob_container_name, self._blob_account_name,
File "/home/cruise/mambaforge/lib/python3.10/site-packages/azureml/opendatasets/dataaccess/_blob_accessor.py", line 304, in
self._get_relative_path(path)) for path in target_paths]
File "/home/cruise/mambaforge/lib/python3.10/site-packages/azureml/opendatasets/dataaccess/_blob_accessor.py", line 470, in _get_relative_path
if "blob.core.windows.net" in url:
TypeError: argument of type 'azureml.dataprep.rslex.StreamInfo' is not iterable

Code to reproduce:

import azureml.core
from azureml.core import Datastore, Dataset
from azureml.core.workspace import Workspace
from azureml.core.experiment import Experiment
from azureml.core.authentication import InteractiveLoginAuthentication

import logging
import pandas as pd
import time
import sys

print("Testing opendatasets -- start")
from azureml.opendatasets import NycTlcYellow
from datetime import datetime
from dateutil import parser
 
end_date = parser.parse('2018-05-30')
start_date = parser.parse('2018-05-28')
nyc_tlc = NycTlcYellow(start_date=start_date, end_date=end_date)


print("Calling to_pandas_dataframe()")
ts = time.time()
nyc_tlc_df = nyc_tlc.to_pandas_dataframe()
te = time.time()
print("Time taken to perform to_pandas_dataframe():" + str(te-ts))

print("Calling to_spark_dataframe()")
ts2 = time.time()
nyc_tlc_df2 = nyc_tlc.to_spark_dataframe()
te2 = time.time()
nyc_tlc_df2.show(2, truncate = False)
print("Time taken to perform to_spark_dataframe():" + str(te2-ts2))

print("Testing opendatasets -- end")