CODEWITHSUNDEEP
pandasmatplotlib
mel
mel

Reputation: 129

2 Plots sharing a common and single legend

I have 2 DFs describing 2 objects with 2 factors they have in common. How do I build 2 charts with a single common legend ?

I try with matplotlib subplots, but i dont' understand how works the legend.

inport pandas as pd

the 1st DF describing Jup :

df1=pd.DataFrame({'Dates' : pd.date_range('2002-02-04',periods=3),
'fac1' : [1,1.02,0.98],
'fac2':[2,2.05,2.06]
 })

the 2d DF describing Tit

df2=pd.DataFrame({'Dates' : pd.date_range('2002-02-04',periods=3),
'fac1' : [1,0.95,1.10],
'fac2':[0.5,0.6,0.55]
})

I can plot this separately like this :

df1.plot(x=df1.iloc[:,0].name,
y=df1.iloc[:,1:3].columns,
legend=True,title='jup')

and this :

df2.plot(x=df2.iloc[:,0].name,
y=df2.iloc[:,1:3].columns,
legend=True,title='tit')

But how do this with a matplotlib style subplots with a common and single legend ?

Upvotes: 1

Views: 5832

Answers (2)

Sheldore
Sheldore

Reputation: 39072

This is an alternate way to do it. The idea is to create two axis objects and then pass them to plot the two DataFrames. Then use legend=False for the first plot as the legends of both the plots are the same. Now you can set the position of the common legend outside the figure using loc.

Complete working answer

import pandas as pd
import matplotlib.pyplot as plt

fig, ax = plt.subplots(nrows=1, ncols=2, figsize=(9, 3))

df1=pd.DataFrame({'Dates' : pd.date_range('2002-02-04',periods=3),
                  'fac1' : [1,1.02,0.98],'fac2':[2,2.05,2.06]})

df2=pd.DataFrame({'Dates' : pd.date_range('2002-02-04',periods=3),
                  'fac1' : [1,0.95,1.10], 'fac2':[0.5,0.6,0.55]}) 

df1.plot(x=df1.iloc[:,0].name, y=df1.iloc[:,1:3].columns, legend=False, title='jup', ax=ax[0])

df2.plot(x=df2.iloc[:,0].name, y=df2.iloc[:,1:3].columns, legend=True, title='tit', ax=ax[1])

ax[1].legend(loc=(1.1, 0.5))
plt.show()

enter image description here

Upvotes: 1

somiandras
somiandras

Reputation: 227

Checkout this recipe from Matplotlib docs. The key is that you need to create your figure and axes first with subplots and add your legend to the figure instead of the axes instances:

import matplotlib.pyplot as plt

fig, (ax1, ax2) = plt.subplots(2, 1) # change layout to (1, 2) for side-by-side plots

l1 = ax1.plot(df1.set_index('Dates'))
l2 = ax2.plot(df2.set_index('Dates'))

fig.legend(l1, ['fac1', 'fac2']) # You only need to add one set of lines and labels

ax1.set_title('jup')
ax2.set_title('tit')

Also you only need to set Dates as index on the dataframes, and pandas and matplotlib will handle the rest nicely.

Upvotes: 0

Related Questions