python - Changing the formatting of a datetime axis in matplotlib

23
votes

I have a series whose index is datetime that I wish to plot. I want to plot the values of the series on the y axis and the index of the series on the x axis. The Series looks as follows:

2014-01-01     7
2014-02-01     8
2014-03-01     9
2014-04-01     8
...

I generate a graph using plt.plot(series.index, series.values). But the graph looks like:

graph

The problem is that I would like to have only year and month. However, the graph contains hours, minutes and seconds. How can I remove them so that I get my desired formatting?

2
pythonpandasdatetimematplotlibpython-datetime
Could you also share your code to generate this graph? It will make for a better solution to you. - Scott Boston
Thanks @ScottBoston, I added the graphing command. - Sheryl
Do you want days? - elPastor
Thanks @pshep123 no, I only want year and month in yyyy-mm format. If pandas can also convert these into e.g. 2016 March, that would be even nicer. - Sheryl

2 Answers

32
votes
# sample data
import numpy as np
import pandas as pd

N = 30
drange = pd.date_range("2014-01", periods=N, freq="MS")
values = {'values':np.random.randint(1,20,size=N)}
df = pd.DataFrame(values, index=drange)

# use formatters to specify major and minor ticks
import matplotlib.pyplot as plt
import matplotlib.dates as mdates

fig, ax = plt.subplots()
ax.plot(df.index, df.values)
ax.set_xticks(df.index)
ax.xaxis.set_major_formatter(mdates.DateFormatter("%Y-%m"))
ax.xaxis.set_minor_formatter(mdates.DateFormatter("%Y-%m"))
_=plt.xticks(rotation=90)

time series plot

15
votes

You can try something like this:

import matplotlib.dates as mdates
import matplotlib.pyplot as plt
df = pd.DataFrame({'values':np.random.randint(0,1000,36)},index=pd.date_range(start='2014-01-01',end='2016-12-31',freq='M'))
fig,ax1 = plt.subplots()
plt.plot(df.index,df.values)
monthyearFmt = mdates.DateFormatter('%Y %B')
ax1.xaxis.set_major_formatter(monthyearFmt)
_ = plt.xticks(rotation=90)

enter image description here