BUG: get UserWarning about FixedFormatter/FixedLocator in series.plot()

[WestXu]

• 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

pd.Series({"20190102": 10, "20190103": 10}).plot()

Problem description

Got a UserWarning:

/home/.../miniconda3/lib/python3.8/site-packages/pandas/plotting/_matplotlib/core.py:1192: UserWarning: FixedFormatter should only be used together with FixedLocator

And to catch the Warning:

import warnings
with warnings.catch_warnings():
    warnings.filterwarnings("error")
    pd.Series({"20190102": 10, "20190103": 10}).plot()

Got

---------------------------------------------------------------------------
UserWarning                               Traceback (most recent call last)
<ipython-input-5-4339a8efd56e> in <module>
      2 with warnings.catch_warnings():
      3     warnings.filterwarnings("error")
----> 4     pd.Series({"20190102": 10, "20190103": 10}).plot()

~/miniconda3/lib/python3.8/site-packages/pandas/plotting/_core.py in __call__(self, *args, **kwargs)
    845                     data.columns = label_name
    846 
--> 847         return plot_backend.plot(data, kind=kind, **kwargs)
    848 
    849     __call__.__doc__ = __doc__

~/miniconda3/lib/python3.8/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

~/miniconda3/lib/python3.8/site-packages/pandas/plotting/_matplotlib/core.py in generate(self)
    268         for ax in self.axes:
    269             self._post_plot_logic_common(ax, self.data)
--> 270             self._post_plot_logic(ax, self.data)
    271 
    272     def _args_adjust(self):

~/miniconda3/lib/python3.8/site-packages/pandas/plotting/_matplotlib/core.py in _post_plot_logic(self, ax, data)
   1190             xticks = ax.get_xticks()
   1191             xticklabels = [get_label(x) for x in xticks]
-> 1192             ax.set_xticklabels(xticklabels)
   1193             ax.xaxis.set_major_locator(FixedLocator(xticks))
   1194 

~/miniconda3/lib/python3.8/site-packages/matplotlib/axes/_base.py in wrapper(self, *args, **kwargs)
     61 
     62         def wrapper(self, *args, **kwargs):
---> 63             return get_method(self)(*args, **kwargs)
     64 
     65         wrapper.__module__ = owner.__module__

~/miniconda3/lib/python3.8/site-packages/matplotlib/cbook/deprecation.py in wrapper(*args, **kwargs)
    449                 "parameter will become keyword-only %(removal)s.",
    450                 name=name, obj_type=f"parameter of {func.__name__}()")
--> 451         return func(*args, **kwargs)
    452 
    453     return wrapper

~/miniconda3/lib/python3.8/site-packages/matplotlib/axis.py in _set_ticklabels(self, labels, fontdict, minor, **kwargs)
   1791         if fontdict is not None:
   1792             kwargs.update(fontdict)
-> 1793         return self.set_ticklabels(labels, minor=minor, **kwargs)
   1794 
   1795     @cbook._make_keyword_only("3.2", "minor")

~/miniconda3/lib/python3.8/site-packages/matplotlib/axis.py in set_ticklabels(self, ticklabels, minor, **kwargs)
   1728             ticks = self.get_minor_ticks(len(locs))
   1729         else:
-> 1730             self.set_major_formatter(formatter)
   1731             locs = self.get_majorticklocs()
   1732             ticks = self.get_major_ticks(len(locs))

~/miniconda3/lib/python3.8/site-packages/matplotlib/axis.py in set_major_formatter(self, formatter)
   1589         formatter : `~matplotlib.ticker.Formatter`, ``str``, or function
   1590         """
-> 1591         self._set_formatter(formatter, self.major)
   1592 
   1593     def set_minor_formatter(self, formatter):

~/miniconda3/lib/python3.8/site-packages/matplotlib/axis.py in _set_formatter(self, formatter, level)
   1619                 and len(formatter.seq) > 0
   1620                 and not isinstance(level.locator, mticker.FixedLocator)):
-> 1621             cbook._warn_external('FixedFormatter should only be used together '
   1622                                  'with FixedLocator')
   1623 

~/miniconda3/lib/python3.8/site-packages/matplotlib/cbook/__init__.py in _warn_external(message, category)
   2088             break
   2089         frame = frame.f_back
-> 2090     warnings.warn(message, category, stacklevel)
   2091 
   2092 

UserWarning: FixedFormatter should only be used together with FixedLocator

Output of pd.show_versions()

INSTALLED VERSIONS

commit : None
python : 3.8.5.final.0
python-bits : 64
OS : Linux
OS-release : 4.19.104-microsoft-standard
machine : x86_64
processor : x86_64
byteorder : little
LC_ALL : None
LANG : C.UTF-8
LOCALE : en_US.UTF-8

pandas : 1.0.5
numpy : 1.19.1
pytz : 2020.1
dateutil : 2.8.1
pip : 20.2.1
setuptools : 49.3.1.post20200810
Cython : None
pytest : 6.0.1
hypothesis : None
sphinx : 3.2.0
blosc : None
feather : None
xlsxwriter : None
lxml.etree : 4.5.2
html5lib : None
pymysql : 0.10.0
psycopg2 : None
jinja2 : 2.11.2
IPython : 7.17.0
pandas_datareader: None
bs4 : 4.9.1
bottleneck : None
fastparquet : None
gcsfs : None
lxml.etree : 4.5.2
matplotlib : 3.3.0
numexpr : 2.7.1
odfpy : None
openpyxl : 3.0.4
pandas_gbq : None
pyarrow : 1.0.0
pytables : None
pytest : 6.0.1
pyxlsb : None
s3fs : None
scipy : 1.5.2
sqlalchemy : 1.3.18
tables : 3.6.1
tabulate : 0.8.7
xarray : None
xlrd : 1.2.0
xlwt : None
xlsxwriter : None
numba : None

[simonjayhawkins]

Thanks @WestXu for the report. This appears to be for Matplotlib 3.3.0+ and persists on pandas master.

>>> pd.__version__
'1.2.0.dev0+100.ged1165575'
>>>
>>> pd.Series({"20190102": 10, "20190103": 10}).plot()
C:\Users\simon\pandas\pandas\plotting\_matplotlib\core.py:1235: UserWarning: FixedFormatter should only be used together with FixedLocator
  ax.set_xticklabels(xticklabels)
<AxesSubplot:>
>>> import matplotlib as mpl
>>> mpl.__version__
'3.3.1'
>>>

>>> import pandas as pd
>>>
>>> pd.__version__
'1.2.0.dev0+100.ged1165575'
>>>
>>> pd.Series({"20190102": 10, "20190103": 10}).plot()
<matplotlib.axes._subplots.AxesSubplot object at 0x0000029E1B3DA040>
>>>
>>> import matplotlib as mpl
>>> mpl.__version__
'3.2.1'
>>>

[pythonic2020]

I got the same error with this code that plots dataframe columns (Series):

fig, ax = plt.subplots(figsize=(8,6))

ax.bar(med2008_sub.index, med2008_sub['Gold'], label='Gold', color='gold')
ax.bar(med2008_sub.index, med2008_sub['Silver'], bottom=med2008_sub['Gold'], color='silver', label='Silver')
ax.bar(med2008_sub.index, med2008_sub['Bronze'], bottom=med2008_sub['Gold'] + med2008_sub['Silver'], 
       color='sandybrown', label='Bronze')
ax.set_xticklabels(med2008_sub.index, rotation=45, ha="right", rotation_mode="anchor")
ax.set_ylabel('Number of Gold medals - 2008')
ax.legend(title='Medal type:')

plt.show()

Result (note error at top):

pd.show_versions() output:

INSTALLED VERSIONS

commit : f2ca0a2
python : 3.7.8.final.0
python-bits : 64
OS : Windows
OS-release : 10
Version : 10.0.19041
machine : AMD64
processor : Intel64 Family 6 Model 26 Stepping 4, GenuineIntel
byteorder : little
LC_ALL : None
LANG : None
LOCALE : None.None

pandas : 1.1.1
numpy : 1.19.1
pytz : 2020.1
dateutil : 2.8.1
pip : 20.2.2
setuptools : 49.6.0.post20200814
Cython : None
pytest : None
hypothesis : None
sphinx : 3.2.1
blosc : None
feather : None
xlsxwriter : None
lxml.etree : 4.5.2
html5lib : None
pymysql : None
psycopg2 : None
jinja2 : 2.11.2
IPython : 7.17.0
pandas_datareader: None
bs4 : 4.9.1
bottleneck : None
fsspec : None
fastparquet : None
gcsfs : None
matplotlib : 3.3.1
numexpr : 2.7.1
odfpy : None
openpyxl : None
pandas_gbq : None
pyarrow : None
pytables : None
pyxlsb : None
s3fs : None
scipy : 1.5.2
sqlalchemy : None
tables : 3.6.1
tabulate : None
xarray : None
xlrd : 1.2.0
xlwt : None
numba : None

​

[CatChenal]

The problem comes from Matplotlib: I can reproduce the problem without using Pandas.

• Info:

Matplotlib: 3.3.1
Python: 3.7.6 | packaged by conda-forge | (default, Jun  1 2020, 18:11:50) [MSC v.1916 64 bit (AMD64)]
Platform: win32

• Test data:

np.random.seed(123)
x = np.array(range(10))
y = np.random.random_sample(10)*4

• Plot triggering the warning:

fig, ax = plt.subplots(1, figsize=(8,5))
ax.bar(x, y);

title1 = "UserWarning: FixedFormatter should only be used together with FixedLocator"
labels = [str(np.round(item.get_position()[1], 3))
              for item in ax.get_yticklabels()]

ax.set(xlim=None,
        xticks=[],
        xlabel='',
        yticklabels=labels,
        ylabel='Series values',
        title=title1)

plt.show();

• Amended code that removes the warning:

fig, ax = plt.subplots(1, figsize=(8,5))
ax.bar(x, y);

title2 = "No UserWarning using set_major_formatter"
ax.yaxis.set_major_formatter('{x:.1f}')
ax.set(xlim=None,
       xticks=[],
       xlabel='',
       ylabel='Series values',
       title=title2
      )
plt.show();

[pythonic2020]

In my example above, if I add

ax.yaxis.set_major_formatter('{x:0.0f}')

the plot is the same as before, but I still get the UserWarning.

[CatChenal]

You did not provide the data in your example, so I cannot reproduce it, but it seems to me that the problem originates with the x axis.

• Instead of

ax.set_xticklabels(med2008_sub.index, rotation=45, ha="right", rotation_mode="anchor")

• Try first:

ax.figure.canvas.draw()
ax.xaxis.set_major_formatter('{x:0.0f}')
ax.tick_params(axis='x',which='major', rotation=45)

• Then:

ax.figure.canvas.draw()
ax.tick_params(axis='x',which='major', rotation=45)

[pythonic2020]

Thanks! Your second idea works, with no Warning (!), but I need to get the

ha="right", rotation_mode="anchor"

code in there somehow so that the x axis labels are placed properly under the bars. I'm still trying various methods...

Is there a matplotlib bug filed for this issue? I couldn't find it...

[CatChenal]

There are many parameters to ax.tick_params: you need to check the documentation then experiment to find out which param can reproduce the previous behavior.

• For help:

ax.tick_params?

• Lastly, have you tried this:

ax.figure.canvas.draw()
ax.set_xticklabels(med2008_sub.index, rotation=45, ha="right", rotation_mode="anchor")

[pythonic2020]

ax.tick_params appears to have no params for anchoring the axis labels. My solution is below. The canvas.draw appears to have no effect.

Your suggestion

ax.figure.canvas.draw()
ax.set_xticklabels(med2008_sub.index, rotation=45, ha="right", rotation_mode="anchor")

produces the correct plot (as my original code did), but this UserWarning is thrown:

C:\Users\Me\Miniconda3\envs\forge37\lib\site-packages\ipykernel_launcher.py:9: UserWarning: FixedFormatter should only be used together with FixedLocator
if name == 'main':

This code works, with no Warning:

fig, ax = plt.subplots(figsize=(8,6))

ax.bar(med2008_sub.index, med2008_sub['Gold'], label='Gold', color='gold')
ax.bar(med2008_sub.index, med2008_sub['Silver'], bottom=med2008_sub['Gold'], color='silver', label='Silver')
ax.bar(med2008_sub.index, med2008_sub['Bronze'], bottom=med2008_sub['Gold'] + med2008_sub['Silver'], 
       color='sandybrown', label='Bronze')

ax.set_xticks(med2008_sub.index)
ax.set_xticklabels(med2008_sub.index, rotation=45, ha="right", rotation_mode="anchor")

ax.set_ylabel('Number of Gold medals - 2008')
ax.legend(title='Medal type:')

plt.show()

All that was needed was the set_xticks.

[simonjayhawkins]

I think this may have been fixed in #35946

code sample from #35684 (comment)

>>> import pandas as pd
>>>
>>> pd.__version__
'1.2.0.dev0+261.g9fea06cec'
>>>
>>> pd.Series({"20190102": 10, "20190103": 10}).plot()
<AxesSubplot:>
>>>
>>> import matplotlib as mpl
>>>
>>> mpl.__version__
'3.3.1'
>>>

maybe should add a release note and close

cc @VirosaLi