[Python] Using DateFormatter resets starting date to 1970

I have a dataframe where the index is the first date of each month and the size column is the frequency for that month, e.g.



Using .index on the dataframe confirms the type of the index is DatetimeIndex:

DatetimeIndex(['2006-12-01', ...],
dtype='datetime64[ns]', name='created_at_month', length=175, freq=None)


Using .plot() on the DataFrame I can produce a line graph per month:



However, it only lists every other year on the x axis, and I'd like it to list each year on the axis.

I would expect to be able to do

ax.xaxis.set_major_locator(mdates.YearLocator(1))
ax.xaxis.set_major_formatter(mdates.DateFormatter('%Y'))


However this doesn't output any labels at all.

If I add a minor formatter (ax.xaxis.set_minor_formatter(mdates.DateFormatter('%d %m %Y'))), I get this:



What am I doing wrong here to cause the dates to change?

The relevant versions are:

• Matplotlib: 3.3.4
• Pandas: 1.2.4
• Python: 3.8.8

Continue reading...