BUG: Offset-based rolling window, with only one raw in dataframe and closed='left', max and min functions make python crash

[jh-wu]

Code Sample

import pandas as pd

# Case 1
df = pd.Series(data=[2], index=pd.date_range('2000', periods=1))
s = df.rolling('10D', closed='left').max()
print(s)

# Case 2 -- by groupby
df = pd.DataFrame(data={'A':[1,1,2],'B': [3,2,1]}, index=pd.date_range('2000', periods=3))
s = df.groupby('A', sort = False)['B'].rolling('10D', closed='left').max()
print(s)

Problem description

This issue is very much similar to issue #21704. So that I copied its problem description below. The only difference is that there is only one entry in the series, which causes the problem.

(From #21704)
"With this rolling and aggregation function, python just crashes. It does too with .min() or .agg(np.max) as aggregation steps. It does not when closed='right' or with mean as an aggregation function."

Expected Output

#Case 1
2000-01-01    NaN
Freq: D, dtype: float64

#Case 2
A            
1  2000-01-01    NaN
   2000-01-02    3.0
2  2000-01-03    NaN
Name: B, dtype: float64

Output of pd.show_versions()

INSTALLED VERSIONS

commit: None
python: 3.7.2.final.0
python-bits: 64
OS: Darwin
OS-release: 18.2.0
machine: x86_64
processor: i386
byteorder: little
LC_ALL: None
LANG: en_AU.UTF-8
LOCALE: en_AU.UTF-8

pandas: 0.24.0.dev0+1563.gd46d2a5ea
pytest: 4.1.0
pip: 18.1
setuptools: 40.6.2
Cython: 0.29.2
numpy: 1.15.4
scipy: 1.2.0
pyarrow: 0.11.1
xarray: 0.11.2
IPython: 7.2.0
sphinx: 1.8.3
patsy: 0.5.1
dateutil: 2.7.5
pytz: 2018.9
blosc: 1.7.0
bottleneck: 1.2.1
tables: 3.4.4
numexpr: 2.6.9
feather: None
matplotlib: 3.0.2
openpyxl: 2.5.12
xlrd: 1.2.0
xlwt: 1.3.0
xlsxwriter: 1.1.2
lxml.etree: 4.3.0
bs4: 4.7.1
html5lib: 1.0.1
sqlalchemy: 1.2.15
pymysql: None
psycopg2: None
jinja2: 2.10
s3fs: 0.2.0
fastparquet: 0.2.1
pandas_gbq: None
pandas_datareader: None
gcsfs: None

[WillAyd]

OK thanks for the report. Investigation and PRs are always welcome!

[jh-wu]

Hi @WillAyd, I have spent some time on investigating this bug today. It seems to me that it is caused by not handling the case when the deque Q is empty in function _roll_min_max_variable of window.pyx (line 1342). I created a PR for this and please review when you have time. Thanks!