In [1]:

    

import datetime

import pandas
import matplotlib.pyplot as plt
import matplotlib.dates


    

In [2]:

    

n = 100
# generate a timeseries of 100 days
dates = [datetime.datetime(2000,10,1) + datetime.timedelta(days=i) for i in range(n)]
values = [x**0.5 for x in range(n)]
df = pandas.DataFrame(data=dict(dates=dates, values=values))
df.head()


    

    
Out[2]:






  

    

      

      
dates
      
values
    
  
  

    

      
0
      
2000-10-01
      
 0.000000
    
    

      
1
      
2000-10-02
      
 1.000000
    
    

      
2
      
2000-10-03
      
 1.414214
    
    

      
3
      
2000-10-04
      
 1.732051
    
    

      
4
      
2000-10-05
      
 2.000000
    
  

5 rows × 2 columns

In [3]:

    

# This is the manual apprach
# create a plot
fig, ax = plt.subplots()
# Create a month locator, creating tick marks at every month
locator = matplotlib.dates.MonthLocator()
# Create a month formatter using year-month notation
formatter = matplotlib.dates.DateFormatter('%Y-%m')
# Add the dates as numbers
ax.plot(matplotlib.dates.date2num(df['dates']), df['values'])
# And format them using the formatter
ax.xaxis.set_major_locator(locator)
ax.xaxis.set_major_formatter(formatter)


    

In [4]:

    

# Or simpler:
df.plot(x='dates', y='values')
# notice the combined week/month grid


    

    
Out[4]:





<matplotlib.axes.AxesSubplot at 0x7f7aa6961ad0>






    

In [5]:

    

# or set the dates as an index
df = df.set_index('dates')
# and then plot the series
df['values'].plot()


    

    
Out[5]:





<matplotlib.axes.AxesSubplot at 0x7f7aa684d090>






    

In [5]:

    




    

In [5]: