BUG: OverflowError when plotting with sharex=True

[shaharkadmiel]

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

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

• (optional) I have confirmed this bug exists on the master branch of pandas.

---

I have two datasets from different sources. They span more or less the same time, contain different data, sampled at different rate.

I am trying to plot these on two subplots with a shared x axis.

Making sample data:

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

time1 = pd.date_range(
    '2020-11-08T09:31:06', '2020-12-01T22:59:59', freq='S', tz='UTC'
)

df1 = pd.DataFrame(
    np.sin(2 * np.pi * 1e-5 * np.arange(time1.size)),
    index=time1,
    columns=['x1']
)

time2 = pd.date_range(
    '2020-11-08T09:32:00', '2020-12-01T22:58:00', freq='D', tz='UTC'
)

df2 = pd.DataFrame(
    np.sin(2 * np.pi * 1e-1 * np.arange(time2.size)),
    index=time2,
    columns=['x2']
)

Minimal plotting example:

fig, (ax1, ax2) = plt.subplots(2, sharex=True)
fig.subplots_adjust(0, 0, 1, 1, hspace=0.15)


df1.x1.plot(ax=ax1, legend=False)

df2.x2.plot(ax=ax2, legend=False)

I get OverflowError and a very long traceback but I think the problem is how the sharex=True is handled. When I set sharex=False, the error goes away and it plots fine but unaligned.

Output of pd.show_versions()

I get ImportError: Can't determine version for pip
But:
Pandas: 1.2.3
Matplotlib: 3.3.2

[paste the output of pd.show_versions() here leaving a blank line after the details tag]

[mzeitlin11]

Hi @shaharkadmiel, thanks for the report! I can reproduce this on current master - based on the traceback it is possible this is a matplotlib issue as well. Investigations welcome!

[shaharkadmiel]

This works:

fig, (ax1, ax2) = plt.subplots(2, sharex=True)
fig.subplots_adjust(0, 0, 1, 1, hspace=0.15)


ax1.plot(df1.index, df1.x1)

ax2.plot(df2.index, df2.x2)

So I am not so sure this is a matplotlib issue.

[konnyakmannan]

Hi, I got the same error and have investigated it.
I suppose the problem is handling sharex=True through matplotlib interface.

works properly with pandas API, like this

fig, (ax1, ax2) = plt.subplots(2, sharex=False)
fig.subplots_adjust(0, 0, 1, 1, hspace=0.15)

df1.x1.plot(ax=ax1, legend=False, sharex=True)

df2.x2.plot(ax=ax2, legend=False, sharex=True)

versions

INSTALLED VERSIONS
------------------
commit           : 2cb96529396d93b46abab7bbc73a208e708c642e
python           : 3.8.5.final.0
python-bits      : 64
OS               : Linux
OS-release       : 5.10.0-1029-oem
Version          : #30-Ubuntu SMP Fri May 28 23:53:50 UTC 2021
machine          : x86_64
processor        : x86_64
byteorder        : little
LC_ALL           : None
LANG             : ja_JP.UTF-8
LOCALE           : ja_JP.UTF-8

pandas           : 1.2.4
numpy            : 1.20.3
pytz             : 2021.1
dateutil         : 2.8.1
pip              : 20.2.2
setuptools       : 49.6.0
Cython           : None
pytest           : None
hypothesis       : None
sphinx           : None
blosc            : None
feather          : None
xlsxwriter       : None
lxml.etree       : None
html5lib         : None
pymysql          : None
psycopg2         : None
jinja2           : None
IPython          : None
pandas_datareader: None
bs4              : None
bottleneck       : None
fsspec           : None
fastparquet      : None
gcsfs            : None
matplotlib       : 3.4.2
numexpr          : None
odfpy            : None
openpyxl         : None
pandas_gbq       : None
pyarrow          : None
pyxlsb           : None
s3fs             : None
scipy            : None
sqlalchemy       : None
tables           : None
tabulate         : None
xarray           : None
xlrd             : None
xlwt             : None
numba            : None
>>> 

[azjps]

This is due to the same underlying issue as #52895