BUG: Series.plot.bar() raises TypeError on timedelta dtype

[darkdragon-001]

• 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.

Code Sample, a copy-pastable example

import pandas as pd
df = pd.DataFrame({'ts1': pd.to_datetime([0, 1, 2, 3], unit='s'), 'ts2': pd.to_datetime([1, 1, 5, 4], unit='s')})
diff = df['ts2']-df['ts1']
diff.plot.bar()

Problem description

Pandas uses fillna() to produce valid data points. Unforunately, the data type timedelta64[ns] doesn't support setting to integer 0.

Output:

TypeError                                 Traceback (most recent call last)
 in 
      2 df = pd.DataFrame({'ts1': pd.to_datetime([0, 1, 2, 3], unit='s'), 'ts2': pd.to_datetime([1, 1, 5, 4], unit='s')})
      3 diff = df['ts2']-df['ts1']
----> 4 diff.plot.bar()
~/.local/lib/python3.9/site-packages/pandas/plotting/_core.py in bar(self, x, y, **kwargs)

1111         other axis represents a measured value.

1112         """

-> 1113         return self(kind="bar", x=x, y=y, **kwargs)

1114

1115     @Appender(
~/.local/lib/python3.9/site-packages/pandas/plotting/_core.py in call(self, *args, **kwargs)

953                     data.columns = label_name

954

--> 955         return plot_backend.plot(data, kind=kind, **kwargs)

956

957     call.doc = doc
~/.local/lib/python3.9/site-packages/pandas/plotting/_matplotlib/init.py in plot(data, kind, **kwargs)

59             kwargs["ax"] = getattr(ax, "left_ax", ax)

60     plot_obj = PLOT_CLASSES[kind](data, **kwargs)

---> 61     plot_obj.generate()

62     plot_obj.draw()

63     return plot_obj.result
~/.local/lib/python3.9/site-packages/pandas/plotting/_matplotlib/core.py in generate(self)

278         self._compute_plot_data()

279         self._setup_subplots()

--> 280         self._make_plot()

281         self._add_table()

282         self._make_legend()
~/.local/lib/python3.9/site-packages/pandas/plotting/_matplotlib/core.py in _make_plot(self)

1419         K = self.nseries

1420

-> 1421         for i, (label, y) in enumerate(self._iter_data(fillna=0)):

1422             ax = self._get_ax(i)

1423             kwds = self.kwds.copy()
~/.local/lib/python3.9/site-packages/pandas/plotting/_matplotlib/core.py in _iter_data(self, data, keep_index, fillna)

256             data = self.data

257         if fillna is not None:

--> 258             data = data.fillna(fillna)

259

260         for col, values in data.items():
~/.local/lib/python3.9/site-packages/pandas/core/frame.py in fillna(self, value, method, axis, inplace, limit, downcast)

4457         downcast=None,

4458     ) -> Optional[DataFrame]:

-> 4459         return super().fillna(

4460             value=value,

4461             method=method,
~/.local/lib/python3.9/site-packages/pandas/core/generic.py in fillna(self, value, method, axis, inplace, limit, downcast)

6424

6425             elif not is_list_like(value):

-> 6426                 new_data = self._mgr.fillna(

6427                     value=value, limit=limit, inplace=inplace, downcast=downcast

6428                 )
~/.local/lib/python3.9/site-packages/pandas/core/internals/managers.py in fillna(self, value, limit, inplace, downcast)

619

620     def fillna(self, value, limit, inplace: bool, downcast) -> "BlockManager":

--> 621         return self.apply(

622             "fillna", value=value, limit=limit, inplace=inplace, downcast=downcast

623         )
~/.local/lib/python3.9/site-packages/pandas/core/internals/managers.py in apply(self, f, align_keys, ignore_failures, **kwargs)

425                     applied = b.apply(f, **kwargs)

426                 else:

--> 427                     applied = getattr(b, f)(**kwargs)

428             except (TypeError, NotImplementedError):

429                 if not ignore_failures:
~/.local/lib/python3.9/site-packages/pandas/core/internals/blocks.py in fillna(self, value, **kwargs)

2478         if is_integer(value):

2479             # Deprecation GH#24694, GH#19233

-> 2480             raise TypeError(

2481                 "Passing integers to fillna for timedelta64[ns] dtype is no "

2482                 "longer supported.  To obtain the old behavior, pass "
TypeError: Passing integers to fillna for timedelta64[ns] dtype is no longer supported.  To obtain the old behavior, pass pd.Timedelta(seconds=n) instead.

Expected Output

Bar plot.

Output of pd.show_versions()

INSTALLED VERSIONS
------------------
commit           : 9d598a5e1eee26df95b3910e3f2934890d062caa
python           : 3.9.1.final.0
python-bits      : 64
OS               : Linux
OS-release       : 5.10.8-200.fc33.x86_64
Version          : #1 SMP Sun Jan 17 19:53:01 UTC 2021
machine          : x86_64
processor        : x86_64
byteorder        : little
LC_ALL           : en_US.UTF-8
LANG             : en_US.UTF-8
LOCALE           : en_US.UTF-8
pandas           : 1.2.1

numpy            : 1.19.5

pytz             : 2020.5

dateutil         : 2.8.1

pip              : 20.1.1

setuptools       : 51.1.0

Cython           : None

pytest           : None

hypothesis       : None

sphinx           : None

blosc            : None

feather          : None

xlsxwriter       : None

lxml.etree       : 4.6.2

html5lib         : 1.1

pymysql          : None

psycopg2         : None

jinja2           : 2.11.2

IPython          : 7.19.0

pandas_datareader: None

bs4              : 4.9.3

bottleneck       : None

fsspec           : None

fastparquet      : None

gcsfs            : None

matplotlib       : 3.3.3

numexpr          : None

odfpy            : None

openpyxl         : None

pandas_gbq       : None

pyarrow          : None

pyxlsb           : None

s3fs             : None

scipy            : None

sqlalchemy       : None

tables           : None

tabulate         : 0.8.7

xarray           : None

xlrd             : None

xlwt             : None

numba            : None

[simonjayhawkins]

Thanks @darkdragon-001 for the report.

This was working in pandas 0.25.3 but was already raising a warning, but will label as regression

C:\Users\simon\anaconda3\envs\pandas-0.25.3\lib\site-packages\pandas\plotting_matplotlib\core.py:253: FutureWarning: Passing integers to fillna is deprecated, will raise a TypeError in a future version. To retain the old behavior, pass pd.Timedelta(seconds=n) instead.
data = data.fillna(fillna)

This warning should not have been visible when using the plot functionality. deprecation in #24694, cc @jbrockmendel

[sopwerdna]

This should probably be slightly higher priority than just in the "contributions welcome" category - basically any attempt to make a bar chart of Timedeltas is broken by this issue.
A much more basic example is the following:

import pandas as pd
pd.DataFrame({'c1': [pd.to_timedelta('1s'), pd.to_timedelta('2s')]}).plot.bar()

The only workaround is to cast your data to timedelta64[s], however this is not documented anywhere and cannot easily be specified as an argument when creating the Timedeltas.
I don't have time to mess with this myself, or I would take care of it, and I certainly will if it is still unfixed during the summer when I have more free time, but I don't think it's something that should go unaddressed for months.
Thanks