load_table_toDataframe breaks with Arrow list fields when the list is backed by a ChunkedArray.

[cvm-a]

Thanks for stopping by to let us know something could be better!

PLEASE READ: If you have a support contract with Google, please create an issue in the support console instead of filing on GitHub. This will ensure a timely response.

Please run down the following list and make sure you've tried the usual "quick fixes":

• Search the issues already opened: https://github.com/googleapis/python-bigquery/issues
• Search StackOverflow: https://stackoverflow.com/questions/tagged/google-cloud-platform+python

If you are still having issues, please be sure to include as much information as possible:

Environment details

• OS type and version: MacOS Darwin Kernel Version 22.1.0
• Python version: 3.11
• pip version: 23.2.1
• google-cloud-bigquery version: 3.15.0

Steps to reproduce

1. Create an Arrow backed dataframe with a large list field.
2. Create a google.cloud.bigquery Client
3. call Client.load_table_from_dataframe on this dataframe

Code example

# example
import pandas as pd
import pyarrow as pa
from google.cloud import bigquery as gbq

client= gbq.Client(
            project=<project_id>,
            credentials=<credentials>,
            location=<location>,
        )
df = pd.DataFrame({"x":pa.array(pd.Series([[2.2]*5]*10000000)).to_pandas(types_mapper=pd.ArrowDtype)})
client.load_table_from_dataframe(
            df, 'temporary_tables.chunked_array_error')

Stack trace

  File "/Users/<redacted>", line 250, in create_table_from_dataframe
    load_job = self.client.load_table_from_dataframe(
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/<redacted>/lib/python3.11/site-packages/google/cloud/bigquery/client.py", line 2671, in load_table_from_dataframe
    new_job_config.schema = _pandas_helpers.dataframe_to_bq_schema(
                            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/<redacted>/lib/python3.11/site-packages/google/cloud/bigquery/_pandas_helpers.py", line 465, in dataframe_to_bq_schema
    bq_schema_out = augment_schema(dataframe, bq_schema_out)
                    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/<redacted>lib/python3.11/site-packages/google/cloud/bigquery/_pandas_helpers.py", line 500, in augment_schema
    arrow_table.values.type.id
    ^^^^^^^^^^^^^^^^^^
AttributeError: 'pyarrow.lib.ChunkedArray' object has no attribute 'values'

Making sure to follow these steps will guarantee the quickest resolution possible.

Thanks!

[cvm-a]

The easiest fix is to replace arrow_table.values.type.id with arrow_table.type.value_type.id.

[Linchin]

I'm unable to reproduce the issue with the code you provided - the BQ table was created successfully. I wonder which version of PyArrow and Pandas you are using? I'm using pyarrow==15.0.0, and pandas==2.2.0. My guess is we don't support converting ChunkedArray into BigQuery types, but your pandas/pyarrow decided to use it.

[Linchin]

I will close this issue now. For future people who are interested in loading an Arrow ChunkedArray, please consider using Arrow's Table.from_arrays() to convert it into a table.