Bar Plots in Seaborn

What is `barplotbarplot` in Seaborn?

Seaborn’s barplotbarplot typically plots:

An aggregate (default is mean)
An uncertainty interval (confidence interval)

This is different from Matplotlib bars where you usually plot raw totals.

Example: mean total bill by day

Seaborn barplot

import seaborn as sns
import matplotlib.pyplot as plt
 
tips = sns.load_dataset("tips")
 
plt.figure(figsize=(7, 4))
sns.barplot(data=tips, x="day", y="total_bill")
plt.title("Mean total bill by day")
plt.tight_layout()
plt.show()

Seaborn barplot

import seaborn as sns
import matplotlib.pyplot as plt
 
tips = sns.load_dataset("tips")
 
plt.figure(figsize=(7, 4))
sns.barplot(data=tips, x="day", y="total_bill")
plt.title("Mean total bill by day")
plt.tight_layout()
plt.show()

Add hue

barplot with hue

import seaborn as sns
import matplotlib.pyplot as plt
 
tips = sns.load_dataset("tips")
 
plt.figure(figsize=(7, 4))
sns.barplot(data=tips, x="day", y="total_bill", hue="sex")
plt.title("Mean total bill by day and sex")
plt.tight_layout()
plt.show()

barplot with hue

import seaborn as sns
import matplotlib.pyplot as plt
 
tips = sns.load_dataset("tips")
 
plt.figure(figsize=(7, 4))
sns.barplot(data=tips, x="day", y="total_bill", hue="sex")
plt.title("Mean total bill by day and sex")
plt.tight_layout()
plt.show()

Change estimator (sum, median, etc.)

Custom estimator

import seaborn as sns
import matplotlib.pyplot as plt
import numpy as np
 
tips = sns.load_dataset("tips")
 
plt.figure(figsize=(7, 4))
sns.barplot(data=tips, x="day", y="total_bill", estimator=np.median)
plt.title("Median total bill by day")
plt.tight_layout()
plt.show()

Custom estimator

import seaborn as sns
import matplotlib.pyplot as plt
import numpy as np
 
tips = sns.load_dataset("tips")
 
plt.figure(figsize=(7, 4))
sns.barplot(data=tips, x="day", y="total_bill", estimator=np.median)
plt.title("Median total bill by day")
plt.tight_layout()
plt.show()

Tips

Use barplotbarplot for comparing group summaries.
Use countplotcountplot when you want raw counts.
If you already aggregated your data with Pandas, consider Matplotlib bars.

If this helped you, consider buying me a coffee ☕

Buy me a coffee