1 seaborn introduction

Introduction to
Seaborn
IN TERMEDIATE DATA VIS UALIZ ATION W ITH S EABORN
Chris Mof tt
Instructor

INTERMEDIATE DATA VISUALIZATION WITH SEABORN
Python Visualization Landscape
The python visualization landscape is complex and can be
overwhelming

Matplotlib
matplotlib provides the raw building blocks for Seaborn's
visualizations
It can also be used on its own to plot data
import matplotlib.pyplot as plt
import pandas as pd
df = pd.read_csv("wines.csv")
fig, ax = plt.subplots()
ax.hist(df['alcohol'])

Pandas
pandas is a foundational library for analyzing data
It also supports basic plotting capability
import pandas as pd
df = pd.read_csv("wines.csv")
df['alcohol'].plot.hist()

Seaborn
Seaborn supports complex visualizations of data
It is built on matplotlib and works best with pandas' dataframes

Seaborn
The distplot is similar to the histogram shown in previous
examples
By default, generates a Gaussian Kernel Density Estimate (KDE)
import seaborn as sns
sns.distplot(df['alcohol'])

Histogram vs. Distplot
Pandas histogram
df['alcohol'].plot.hist()
Actual frequency of
observations
No automatic labels
Wide bins
Seaborn distplot
sns.distplot(df['alcohol'])
Automatic label on x axis
Muted color palette
KDE plot
Narrow bins

Let's practice!

Using the
distribution plot
Chris Mof tt
Instructor

Creating a histogram
Distplot function has multiple optional arguments
In order to plot a simple histogram, you can disable the kde and
specify the number of bins to use
sns.distplot(df['alcohol'], kde=False, bins=10)

Alternative data distributions
A rug plot is an alternative way to view the distribution of data
A kde curve and rug plot can be combined
sns.distplot(df_wines['alcohol'], hist=False, rug=True)

Further Customizations
The distplot function uses several functions including
kdeplot and rugplot
It is possible to further customize a plot by passing arguments to
the underlying function
sns.distplot(df_wines['alcohol'], hist=False,
rug=True, kde_kws={'shade':True})

Regression Plots in
Seaborn
Chris Mof tt
Instructor

Introduction to regplot
The regplot function generates a scatter plot with a regression
line
Usage is similar to the distplot
The data and x and y variables must be de ned
sns.regplot(x="alcohol", y="pH", data=df)

lmplot() builds on top of the base regplot()
regplot - low level
sns.regplot(x="alcohol",
y="quality",
data=df)
lmplot - high level
sns.lmplot(x="alcohol",
y="quality",
data=df)

lmplot faceting
Organize data by colors (
hue )
sns.lmplot(x="quality",
y="alcohol",
data=df,
hue="type")
Organize data by columns (
col )
sns.lmplot(x="quality",
y="alcohol",
data=df,
col="type")

1 seaborn introduction

More Related Content

What's hot (20)

Similar to 1 seaborn introduction (20)

Recently uploaded (20)

1 seaborn introduction