Seaborn Objects

a advanced graphic grammar

Zongzhou Wu
Oct 28, 2021 2 min read research

Introduction

Seaborn is a Python data visualization library based on matplotlib. It provides a high-level interface for drawing attractive and informative statistical graphics.

Seaborn website: https://seaborn.pydata.org/

Seaborn Github: https://github.com/mwaskom/seaborn

Dependencies

Seaborn supports Python 3.8+.

Installation requires numpy, pandas, and matplotlib. Some advanced statistical functionality requires scipy and/or statsmodels.

Installation

pip install seaborn

Object Interface

In version 0.12, Seaborn introduced object namespace and interface, which are based on the Grammar of Graphics similiar to ggplot2 package in R. In the past, the best python package was plotnine that was absolutely consistent with ggplot2.

Example

Basic Plot

import seaborn
import seaborn.objects as so

tips=seaborn.load_dataset("tips")

(
    so.Plot(tips, y="day", color="sex")
    .add(so.Bar(), so.Hist(), so.Dodge())
    .show()
)

Then figure will be like this

If you want to save the figure, just run

import seaborn as sns
import seaborn.objects as so

tips=sns.load_dataset("tips")

(
    so.Plot(tips, y="day", color="sex")
    .add(so.Bar(), so.Hist(), so.Dodge())
    .save("fig.pdf",bbox_inches="tight") ## save
)

bbox_inches=“tight” is necessary for saving legend like R.

Facet

import seaborn as sns
import seaborn.objects as so

penguins=sns.load_dataset("penguins")

(
    so.Plot(penguins, "bill_length_mm", "bill_depth_mm")
    .add(so.Dots())
    .facet("species", "sex")
    .show()
)

Theme and Palette

import seaborn as sns
import seaborn.objects as so
import matplotlib.pyplot as plt

sns.set_theme(style="white", palette="deep6", font="Times New Roman", font_scale=1.5) ## settings

tips=sns.load_dataset("tips")

(
    so.Plot(tips, "total_bill", "tip", color="day")
    .facet(col="day")
    .add(so.Dot(color="#aabc"), col=None, color=None)
    .add(so.Dot())
    .theme(plt.rcParams) ## important
    .show()
)

Reference

https://seaborn.pydata.org/api.html#objects-api

https://seaborn.pydata.org/tutorial/objects_interface.html

tool
Zongzhou Wu
Zongzhou Wu
Ph.D. Student

My research interest includes multi-sensor fusion, Global Navigation Satellite System (GNSS), indoor-outdoor seamless positioning, simultaneous localization and mapping (SLAM), and sensor calibration.