14

I want to plot two pandas dataframes side by side, each plot should be in subplot form. I am using following lines:

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
# create dummy pandas dataframes
pd1 = pd.DataFrame({'a':np.random.random(22),'b':np.random.random(22),'c':np.random.random(22)})
pd2 = pd.DataFrame({'J':np.random.random(22),'K':np.random.random(22),'P':np.random.random(22)})
#create subplot figure with having two side by side plots
fig, axes = plt.subplots(nrows=1,ncols=2,figsize=(12,6))
# plot first pandas frame in subplot style
pd1.plot(ax = axes[0],subplots=True) 
# plot second pandas frame in subplot style
pd2.plot(ax = axes[1],subplots=True)

Without subplot option, side by side plots are drawn but I want in subplot style. Is there any other option to get it done?

Precisely, I want to plot pandas in the following manner: enter image description here

Haroon Lone
  • 2,837
  • 5
  • 29
  • 65

1 Answers1

30

You need a subplot grid of 3 rows and 2 columns to host your plot. Then the ax argument needs to take the 3 axes you want to plot the 3 dataframe columns to.

fig, axes = plt.subplots(nrows=3,ncols=2)
pd1.plot(ax = axes[:,0], subplots=True) 
pd2.plot(ax = axes[:,1], subplots=True)

Complete code:

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
# create dummy pandas dataframes
pd1 = pd.DataFrame({'a':np.random.random(22),'b':np.random.random(22),
                    'c':np.random.random(22)})
pd2 = pd.DataFrame({'J':np.random.random(22),'K':np.random.random(22),
                    'P':np.random.random(22)})
#create subplot figure with having two side by side plots
fig, axes = plt.subplots(nrows=3,ncols=2,figsize=(12,6))
# plot first pandas frame in subplot style
pd1.plot(ax = axes[:,0],subplots=True) 
# plot second pandas frame in subplot style
pd2.plot(ax = axes[:,1],subplots=True)

plt.show()

enter image description here

ImportanceOfBeingErnest
  • 321,279
  • 53
  • 665
  • 712
  • I still face a problem when the two data frames have a different number of columns. I think that is a different question and I need to find some workaround for that. – Haroon Lone Feb 27 '18 at 11:49
  • 1
    You need to create the subplot grid such that it can host the number of subplots created by the dataframe plot. Then make sure the list of axes supplied to `ax` has exactly the number of axes as there are columns to plot. Depending on the number of columns to plot you may not be able to use `plt.subplots` to create the grid, but rather create the subplots one by one or using a [`gridspec`](https://matplotlib.org/users/gridspec.html). – ImportanceOfBeingErnest Feb 27 '18 at 11:57