Set DateTimeIndex format when using TimeGroupper

Question

In the following snippet I want to do the format transformation inplace without having to keep another temporary variable (df)

df = other_df.set_index('datetime_created').groupby(pd.TimeGrouper("M")).aggregate({...complex dict...})
df.index = df.index.format(formatter=lambda x: x.strftime('%b %Y'))
df.plot.bar(sharex=True, stacked=True)

First, biggest question: Why? There is no shame in using a temporary variable. — cwallenpoole, Dec 02 '16 at 13:50
When prototyping & plotting in Jupyter I prefer to work with immutables — Flo, Dec 02 '16 at 14:17

score 0 · Answer 1 · answered Dec 02 '16 at 16:41

I guess you could just format the axis after plotting? One example:

df = pd.DataFrame(np.random.rand(10,2), 
                 index=pd.date_range('2005-01-01', freq='M', periods=10))
ax = df.plot(kind='bar', sharex=True, stacked=True)
ax.set_xticklabels([x.strftime('%b %Y') for x in df.index])
plt.show()

Set DateTimeIndex format when using TimeGroupper

1 Answers1