BUG: doing df.to_parquet and then reading it with pd.read_parquet causes KeyError

[elbg]

Pandas version checks

• I have checked that this issue has not already been reported.

• I have confirmed this bug exists on the latest version of pandas.

• I have confirmed this bug exists on the main branch of pandas.

Reproducible Example

import pandas as pd

df = pd.DataFrame(
    {
        "model": ["model1", "model2"],
        "second_index": [(1, 2), (3, 4)],
        "first_index": [0, 1],
    }
)

df = df.set_index(["first_index", "second_index"], append=True)
df.to_parquet("temp.parquet")
pd.read_parquet("temp.parquet") # >> KeyError

import polars as pl
pl.read_parquet('temp.parquet') #--> OK

Issue Description

I am writing a dataframe with a multiindex, some level of the multiindex contains tuples.
I can save it to parquet, and the obtained parquet seems to be valid since polars reads it correctly.

I can't load it back to pandas, it produces a key error.

Expected Behavior

I expected pd.read_parquet to give batck the df from df.to_parquet. This produces the correct result:

df = pl.read_parquet("temp.parquet").to_pandas()
df["second_index"] = df["second_index"].apply(lambda x: tuple(x))
df = df.set_index(["first_index", "second_index"])

Installed Versions

INSTALLED VERSIONS

commit : 2cc3762
python : 3.12.10
python-bits : 64
OS : Linux
OS-release : 5.15.0-139-generic
Version : #149-Ubuntu SMP Fri Apr 11 22:06:13 UTC 2025
machine : x86_64
processor : x86_64
byteorder : little
LC_ALL : None
LANG : C.UTF-8
LOCALE : C.UTF-8

pandas : 2.3.0
numpy : 1.26.4
pytz : 2025.2
dateutil : 2.9.0.post0
pip : 25.1.1
Cython : 3.1.2
sphinx : None
IPython : 9.3.0
adbc-driver-postgresql: None
adbc-driver-sqlite : None
bs4 : 4.13.4
blosc : None
bottleneck : None
dataframe-api-compat : None
fastparquet : 2024.11.0
fsspec : 2024.9.0
html5lib : None
hypothesis : None
gcsfs : None
jinja2 : 3.1.6
lxml.etree : 5.4.0
matplotlib : 3.10.3
numba : None
numexpr : None
odfpy : None
openpyxl : None
pandas_gbq : None
psycopg2 : None
pymysql : None
pyarrow : 20.0.0
pyreadstat : None
pytest : 7.4.4
python-calamine : None
pyxlsb : None
s3fs : None
scipy : 1.16.0
sqlalchemy : None
tables : None
tabulate : 0.9.0
xarray : None
xlrd : None
xlsxwriter : None
zstandard : None
tzdata : 2025.2
qtpy : None
pyqt5 : None

[Jopestpe]

replace

df.to_parquet("temp.parquet")

with

df.reset_index().to_parquet("temp.parquet")

it worked for me, maybe it works for you too.

[elbg]

Yup that's another workaround. However, I would rather have pandas handling the index, rather than resetting it when saving to parquet and setting it back again when loading the parquet

[rhshadrach]

Thanks for the report. Further investigations welcome. This may also be an upstream issue in PyArrow.