[BUG] to_json ignores index=True

[dagardner-nv]

Describe the bug
Documentation for the index argument for cudf.DataFrame.to_json https://docs.rapids.ai/api/cudf/stable/api_docs/api/cudf.DataFrame.to_json.html
states:

indexbool, default True
Whether to include the index values in the JSON string. Not including the index (index=False) is only supported when orient is ‘split’ or ‘table’.

However this is ignored, and appears to be caused by Pandas as cudf's to_json uses Panda's to_json.
pandas-dev/pandas#37600

Steps/Code to reproduce bug
Issue can be reproduced in both cudf and pandas.

cudf repro:

import cudf

df = cudf.DataFrame([3,4,5,6])

# Including CSV output for comparison
print(df.to_csv(header=True, index=True))

print(df.to_json(index=True, orient="records"))

Yields this output:

,0
0,3
1,4
2,5
3,6

[{"0":3},{"0":4},{"0":5},{"0":6}]

Pandas repro:

import pandas

df = pandas.DataFrame([3,4,5,6])

# Including CSV output for comparison
print(df.to_csv(header=True, index=True))

print(df.to_json(index=True, orient="records"))

Expected behavior
Include index column when index=True

Environment overview (please complete the following information)

• Environment location: Bare-metal
• Method of cuDF install: conda

[github-actions]

This issue has been labeled inactive-30d due to no recent activity in the past 30 days. Please close this issue if no further response or action is needed. Otherwise, please respond with a comment indicating any updates or changes to the original issue and/or confirm this issue still needs to be addressed. This issue will be labeled inactive-90d if there is no activity in the next 60 days.

[GregoryKimball]

Thank you for raising this issue. cudf.DataFrame.to_json currently uses a host-fallback to pandas, and it seem that the index argument is not correctly passed.

Seems like this could be related to other changes in #11780

[galipremsagar]

@dagardner-nv

This is expected behavior in pandas to_json, which is documented here: https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.to_json.html

However if you really need index data in json, you will need to pass orient='index' instead, which is shown in the example right below:

Both these example have been snapshotted from the to_json docs page

Hence closing this issue on cudf side as there is no action item.