Skip to content

PERF: Plots with some DatetimeIndex frequencies very slow #50355

Open
@brianbienvenu

Description

@brianbienvenu

Pandas version checks

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

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

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

Reproducible Example

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

for freq in ['1ms', '10ms', '100ms', '1000ms', '1s', '10s', '1min', '100s', '10min', '1000s']:
    for periods in [1e4, 1e5]:
        dates = pd.date_range('2022-01-01 00:00:00', periods=periods, freq=freq)
        values = np.random.rand(len(dates))
        df = pd.DataFrame(data=values, index=dates)

        start = datetime.now()
        df.plot()
        finish = datetime.now()
        print(f'{periods:6.0f} at {freq:<6} took {finish - start}')
        plt.close()

Installed Versions

INSTALLED VERSIONS

commit : 8dab54d
python : 3.11.1.final.0
python-bits : 64
OS : Windows
OS-release : 10
Version : 10.0.22621
machine : AMD64
processor : AMD64 Family 23 Model 113 Stepping 0, AuthenticAMD
byteorder : little
LC_ALL : None
LANG : en_US.UTF-8
LOCALE : English_Australia.1252

pandas : 1.5.2
numpy : 1.23.4
pytz : 2022.6
dateutil : 2.8.2
setuptools : 65.5.0
pip : 22.3.1
Cython : None
pytest : None
hypothesis : None
sphinx : None
blosc : None
feather : None
xlsxwriter : None
lxml.etree : None
html5lib : None
pymysql : None
psycopg2 : None
jinja2 : 3.1.2
IPython : None
pandas_datareader: None
bs4 : None
bottleneck : None
brotli : None
fastparquet : None
fsspec : None
gcsfs : None
matplotlib : 3.6.2
numba : None
numexpr : None
odfpy : None
openpyxl : 3.0.10
pandas_gbq : None
pyarrow : None
pyreadstat : None
pyxlsb : None
s3fs : None
scipy : 1.9.3
snappy : None
sqlalchemy : None
tables : None
tabulate : None
xarray : None
xlrd : None
xlwt : None
zstandard : None
tzdata : None

Prior Performance

I think this has been a long standing issue. Runnig the above code outputs the following:

 10000 at 1ms    took 0:00:00.185187
100000 at 1ms    took 0:00:00.319529
 10000 at 10ms   took 0:00:00.165032
100000 at 10ms   took 0:00:00.933138
 10000 at 100ms  took 0:00:00.793506
100000 at 100ms  took 0:00:07.192190
 10000 at 1000ms took 0:00:07.064933
100000 at 1000ms took 0:00:58.392849
 10000 at 1s     took 0:00:00.226735
100000 at 1s     took 0:00:00.293485
 10000 at 10s    took 0:00:00.144515
100000 at 10s    took 0:00:00.671181
 10000 at 1min   took 0:00:00.093631
100000 at 1min   took 0:00:00.317967
 10000 at 100s   took 0:00:00.532972
100000 at 100s   took 0:00:04.810899
 10000 at 10min  took 0:00:00.277081
100000 at 10min  took 0:00:00.674374
 10000 at 1000s  took 0:00:04.690483
100000 at 1000s  took 0:00:33.029342

There's no easily discernable reason why frequencies of '1000ms' has >100x plot render times than '1s'. The output plots look quite similar.

Metadata

Metadata

Assignees

No one assigned

    Labels

    DatetimeDatetime data dtypePerformanceMemory or execution speed performanceVisualizationplotting

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions