The setting xrot=0 in DataFrame.hist() doesn't work with by and subplots

[PavloFesenko]

Code Sample, a copy-pastable example if possible

df = pd.DataFrame({
    'length': [1.5, 0.5, 1.2, 0.9, 3],
    'animal': ['pig', 'rabbit', 'pig', 'pig', 'rabbit']
})
df.hist(column='length', by='animal', bins=5, xrot=0);

Problem description

The setting xrot=0 doesn't work when I use the option by in DataFrame.hist() (see the screenshot below). The temporary workaround is to set it to a very small value like xrot=0.001.

Expected Output

Should look like this instead

Output of pd.show_versions()

INSTALLED VERSIONS

commit : None
python : 3.6.9.final.0
python-bits : 64
OS : Linux
OS-release : 3.10.0-1062.4.1.el7.x86_64
machine : x86_64
processor : x86_64
byteorder : little
LC_ALL : None
LANG : en_US.UTF-8
LOCALE : en_US.UTF-8

pandas : 0.25.2
numpy : 1.17.2
pytz : 2019.3
dateutil : 2.8.0
pip : 19.3.1
setuptools : 41.4.0
Cython : None
pytest : None
hypothesis : None
sphinx : None
blosc : None
feather : None
xlsxwriter : None
lxml.etree : None
html5lib : None
pymysql : None
psycopg2 : None
jinja2 : 2.10.3
IPython : 7.8.0
pandas_datareader: None
bs4 : None
bottleneck : None
fastparquet : None
gcsfs : None
lxml.etree : None
matplotlib : 3.1.1
numexpr : None
odfpy : None
openpyxl : 3.0.2
pandas_gbq : None
pyarrow : None
pytables : None
s3fs : None
scipy : 1.3.1
sqlalchemy : None
tables : None
xarray : None
xlrd : 1.2.0
xlwt : None
xlsxwriter : None

[TomAugspurger]

pandas/pandas/plotting/_matplotlib/hist.py

Line 253 in 0cd388f

xrot = xrot or rot

should probably be

if xrot is None:
    xrot = rot

since 0 evaluates to False. A PR fixing that + adding a test would be welcome.

[souvik3333]

Hi, Can I do it?

[souvik3333]

I am taking this as no one seems to be working on this.

[souvik3333]

take

[souvik3333]

should probably be

if xrot is None:
    xrot = rot

since 0 evaluates to False. A PR fixing that + adding a test would be welcome.

Looks like

pandas/pandas/plotting/_matplotlib/core.py

Line 1208 in f505b74

self.rot = 30

is causing the error. Assigning self.rot=0 solves the problem. I will add a PR on this. Also @TomAugspurger the solution you said fixes another issue. Should I add that in the same PR or create another issue on this?

[charlesdong1991]

thanks for working on it!
@souvik3333 i think the error is not caused by self.rot=30, this is only used for plotting ts data

and i do think if you follow @TomAugspurger 's suggestion, should be able to fix the bug.

[souvik3333]

i do think if you follow @TomAugspurger 's suggestion, should be able to fix the bug.

I tried that but it's not working. Looks like

pandas/pandas/plotting/_matplotlib/hist.py

Line 250 in 0cd388f

def plot_group(group, ax):

function is not even getting called during the execution of the example given above.

[souvik3333]

I have my doubt but I think the following is the problem.
In dataframe.plot if we have either of subplots=True or sharex=True then it's treated as irregular TS object because of this condition:

pandas/pandas/plotting/_matplotlib/core.py

Line 1199 in f505b74

or (self.subplots and self.sharex)

So, when rot argument is not present in dataframe.plot then it is rotated 30deg. So should I update the condition @TomAugspurger @charlesdong1991 ?

[charlesdong1991]

emm, okay, it seems indeed if both subplots and sharex are True, then will rot to 30. And this should be designed on purpose.

Regarding your question: I did not look at the codebase, but my assumption is when using hist and use by to plot, then subplots and sharex are automatically set to True.

Just follow your investigation, and you could either submit a PR or open a new issue if you think there is another bug. @souvik3333

[souvik3333]

Oh, actually this bug will be fixed as @TomAugspurger said. I misplaced it with another bug I found. My apologies. I will add a new issue on that bug and add a PR on this shortly.