How to plot a grouped bar plot from two or more dataframes

Question

I have multiple dataframes, and I want to plot them on the same figure in the Grouped Bar Chart view.

These are two very small dataframes that I would like to plot together in the same figure.

The dataframes are:

my dataframes

I want to plot a figure like this example:

example of what i want

I try this, plot only one graph:

fig, ax = plt.subplots()

df1.plot.bar(x='Zona',y='Total_MSP')
df4.plot.bar(x='Zona',y='NumEstCasasFavelas2017',ax=ax)

plt.show()

I tried this too:

fig, ax = plt.subplots()

df1.plot.bar(x='Zona',y='Total_MSP',ax=ax)
df4.plot.bar(x='Zona',y='NumEstCasasFavelas2017',ax=ax)

plt.show()

The results are just data from a single dataframe in a picture, not two data from both dataframes. Note that only the captions of both dataframes appear in the same picture, the data is only from a single isolated dataframe.

my failures

Asked By: Gizelly

||

Source

Answer 1

In order to create a grouped bar plot, the DataFrames must be combined with pandas.merge or pandas.DataFrame.merge.
See pandas User Guide: Merge, join, concatenate and compare and SO: Pandas Merging 101.

Data:

import pandas as pd
import matplotlib.pyplot as plt

df1 = pd.DataFrame({'Zone': ['C', 'L', 'N', 'O', 'S'],
                    'Total_MSP': [464245, 3764942, 1877505, 1023160, 3179477]})
df2 = pd.DataFrame({'Zone': ['C', 'L', 'N', 'O', 'S'],
                    'CasasFavelas_2017': [463, 4228, 851, 1802, 2060]})

Merge the dataframes:

Using pandas.merge, combine the DataFrames.

df = pd.merge(df1, df2, on='Zone')

  Zone  Total_MSP  CasasFavelas_2017
0    C     464245                463
1    L    3764942               4228
2    N    1877505                851
3    O    1023160               1802
4    S    3179477               2060

Plot:

Plot the DataFrame with pandas.DataFrame.plot.
- Use log scale for Casas to show up.

ax = df.plot(kind='bar', x='Zone', logy=True, rot=0)
ax.legend(bbox_to_anchor=(1.05, 1), loc='upper left')
plt.show()

Update:

The OP added additional data in an answer, after this answer was provided.
Use pandas.concat to combine more than 2 DataFrames.

df12 = pd.DataFrame({'Zone': ['C', 'L', 'N', 'O', 'S'], 'Total_MSP': [464245, 3764942, 1877505, 1023160, 3179477]})
df13 = pd.DataFrame({'Zone': ['C', 'L', 'N', 'O', 'S'], 'ValorMedioDollar': [1852.27, 1291.53, 1603.44, 2095.90, 1990.10]})
df14 = pd.DataFrame({'Zone': ['C', 'L', 'N', 'O', 'S'], 'IDH2010': [0.89, 0.70, 0.79, 0.90, 0.80]})
df15 = pd.DataFrame({'Zone': ['C', 'L', 'N', 'O', 'S'], 'QtdNovasCasas': [96,1387, 561, 281, 416]})

# use concat to combine more than two DataFrames
df = pd.concat([df12.set_index('Zone'), df13.set_index('Zone'), df14.set_index('Zone'), df15.set_index('Zone')], axis=1)

      Total_MSP  ValorMedioDollar  IDH2010  QtdNovasCasas
Zone                                                     
C        464245           1852.27     0.89             96
L       3764942           1291.53     0.70           1387
N       1877505           1603.44     0.79            561
O       1023160           2095.90     0.90            281
S       3179477           1990.10     0.80            416

# plot the DataFrame
ax = df.plot(kind='bar', logy=True, figsize=(8, 6), rot=0)
ax.legend(bbox_to_anchor=(1.05, 1), loc='upper left')
plt.show()

Adding Annotations:

Not part of the original question.

Answered By: Trenton McKinney

Answer 2

Graphic with four custom color dataframes and caption

import pandas as pd


df12 = pd.DataFrame({'Zone': ['C', 'L', 'N', 'O', 'S'],
                    'Total_MSP': [464245, 3764942, 1877505, 1023160, 3179477]})
df13 = pd.DataFrame({'Zone': ['C', 'L', 'N', 'O', 'S'],
                    'ValorMedioDollar': [1852.27, 1291.53, 1603.44, 2095.90, 1990.10]})
df14 = pd.DataFrame({'Zone': ['C', 'L', 'N', 'O', 'S'],
                    'IDH2010': [0.89, 0.70, 0.79, 0.90, 0.80]})
df15 = pd.DataFrame({'Zone': ['C', 'L', 'N', 'O', 'S'],
                    'QtdNovasCasas': [96,1387, 561, 281, 416]})


df16 = pd.merge(df12, df13, on='Zone')
df16 = pd.merge(df16, df14, on='Zone')
df16 = pd.merge(df16, df15, on='Zone')

fig, ax = plt.subplots(figsize=(50, 20))

#https://xkcd.com/color/rgb/
colors2 = ['#448ee4', '#a9f971','#ceb301','#ffb7ce']


#For all values to be displayed, even though these scales are different, the log scale is used.
df16.plot.bar(x='Zone', logy=True, color=colors2, ax=ax,width=0.5, align = 'center'); 


#legend
#https://stackoverflow.com/questions/19125722/adding-a-legend-to-pyplot-in-matplotlib-in-the-most-simple-manner-possible
plt.gca().legend(('Total Resident Population-2017', 
                  'Median Value of square meter-Dollars US', 
                  'HDI- Human Development Index-2010',
                  'Number of new housing properties-2018'),bbox_to_anchor=(0.87, 0.89) ,fontsize=28)


plt.title('Estimated Resident Population, Average value of square meter, HDI, New housing properties in São Paulo - Brazil',fontsize=40)
plt.xlabel ('Names of the geographical subdivisions of São Paulo',fontsize=40)
plt.ylabel('Log Scale', fontsize=30)

#change the name of month on the x 
ax = plt.gca()
names = ['Zone: Center', 'Zone: East', 'Zone: North', 'Zone: West', 'Zone: South']
ax.set_xticklabels(names,fontsize=40)
x = plt.gca().xaxis



plt.rcParams['ytick.labelsize'] = 30

# rotate the tick labels for the x axis
for item in x.get_ticklabels():
    item.set_rotation(0)    

for spine in plt.gca().spines.values():
    spine.set_visible(False)

# remove all the ticks (both axes), and tick labels on the Y axis
plt.tick_params(top='off', bottom='off', left='off', right='off', labelleft='on', labelbottom='on')



# direct label each bar with Y axis values
for p in ax.patches[0:]:
    plt.gca().text(p.get_x() + p.get_width()/2, p.get_height()+0.01, str(float(p.get_height())), 
                 ha='center', va='baseline', rotation=0 ,color='black', fontsize=25)



plt.show()


fig.savefig('GraficoMultiplo.jpg')

Answered By: Gizelly