Plot a side by side bar charts with sharing only Y-axis

user_11157647 0
Viewed 389

How to plot a side by side bar charts from multiple dataframe with the same Y-axis?

from matplotlib import pyplot as plt
import pandas as pd

fruits = {"Apple": 10, "Banana": 9, "Mango": 5, "Orange": 1}
beverages = {"Coca-cola": 15, "Pepsi": 12, "Seven-up": 9, "Root-bear": 3}
foods = {'hamburger': 10, 'pizza': 7, 'pasta': 5, 'sushi': 2}

fruits_df = pd.DataFrame(fruits.items(), columns = ['stuffs', 'freq'])
beverages_df = pd.DataFrame(beverages.items(), columns = ['stuffs', 'freq'])
foods_df = pd.DataFrame(foods.items(), columns = ['stuffs', 'freq'])

# initialize a figure
fig=plt.figure(figsize=(18, 6))
plt.subplots_adjust(hspace = 2)

ax1 = fig.add_subplot(221)
fruits_df.head(8).plot(x='stuffs', y='freq', kind='bar', ax=ax1, legend=False).set_title('fruits')
plt.xticks(rotation=45, ha='right')

ax2 = fig.add_subplot(222)
beverages_df.head(8).plot(x='stuffs', y='freq', kind='bar', ax=ax2, legend=False, color="orange").set_title('beverage')
plt.xticks(rotation=45, ha='right')

ax2 = fig.add_subplot(223)
foods_df.head(8).plot(x='stuffs', y='freq', kind='bar', ax=ax2, legend=False, color="blue").set_title('foods')
plt.xticks(rotation=45, ha='right')

enter image description here

My desired output is following. I'm planning to add three or more dataframes on it like snacks, costumes and etc... please help

enter image description here

pandas seaborn dataframe matplotlib python
2 Answers

Using plt.subplots(..., sharey=True) might do the trick.

from matplotlib import pyplot as plt
import pandas as pd

fruits = {"Apple": 10, "Banana": 9, "Mango": 5, "Orange": 1}
beverages = {"Coca-cola": 15, "Pepsi": 12, "Seven-up": 9, "Root-bear": 3}
foods = {'hamburger': 10, 'pizza': 7, 'pasta': 5, 'sushi': 2}

fruits_df = pd.DataFrame(fruits.items(), columns = ['stuffs', 'freq'])
beverages_df = pd.DataFrame(beverages.items(), columns = ['stuffs', 'freq'])
foods_df = pd.DataFrame(foods.items(), columns = ['stuffs', 'freq'])

# initialize a figure
fig, axes = plt.subplots(1, 3, figsize=(18, 6), sharey=True)
ax1, ax2, ax3 = axes
fruits_df.plot(x='stuffs', y='freq', kind='bar', ax=ax1, legend=False).set_title('fruits')
beverages_df.plot(x='stuffs', y='freq', kind='bar', ax=ax2, legend=False, color="orange").set_title('beverage')
foods_df.plot(x='stuffs', y='freq', kind='bar', ax=ax3, legend=False, color="blue").set_title('foods')

for ax in axes:
    ax.set_xticklabels(ax.get_xticklabels(), rotation = 45)
    for side in ('right', 'top', 'left'):
        if (ax == ax1) and (side == 'left'):
            continue
        else:
            sp = ax.spines[side]
            sp.set_visible(False)

enter image description here

user_170256880

You can concatenate all dataframe together through:

fruits_df['type'] = 'fruit'
beverages_df['type'] = 'beverages'
foods_df['type'] = 'foods'

df = pd.concat([fruits_df, beverages_df, foods_df], ignore_index = True)

and then you can use seaborn.barplot and pass the parameter hue = 'type':

fig, ax = plt.subplots(figsize = (18, 6))

sns.barplot(ax = ax, data = df, x = 'stuffs', y = 'freq', hue = 'type', dodge = False)

plt.show()

Complete Code

from matplotlib import pyplot as plt
import pandas as pd
import seaborn as sns


fruits = {"Apple": 10, "Banana": 9, "Mango": 5, "Orange": 1}
beverages = {"Coca-cola": 15, "Pepsi": 12, "Seven-up": 9, "Root-bear": 3}
foods = {'hamburger': 10, 'pizza': 7, 'pasta': 5, 'sushi': 2}

fruits_df = pd.DataFrame(fruits.items(), columns = ['stuffs', 'freq'])
beverages_df = pd.DataFrame(beverages.items(), columns = ['stuffs', 'freq'])
foods_df = pd.DataFrame(foods.items(), columns = ['stuffs', 'freq'])

fruits_df['type'] = 'fruit'
beverages_df['type'] = 'beverages'
foods_df['type'] = 'foods'

df = pd.concat([fruits_df, beverages_df, foods_df], ignore_index = True)


fig, ax = plt.subplots(figsize = (18, 6))

sns.barplot(ax = ax, data = df, x = 'stuffs', y = 'freq', hue = 'type', dodge = False)

plt.show()

enter image description here

user_92143570
Related