Seaborn Chart

Code

library(reticulate)
py_require(c('celluloid','seaborn','IPython'))

1 Introduction to Seaborn

Seaborn is a Python data visualization library based on Matplotlib. It provides a high-level interface for creating attractive and informative statistical graphics. This document demonstrates how to create various plots using Seaborn, customize their appearance, and save them.

Code

import seaborn as sns
print(sns.__version__)

0.13.2

Code

# Import necessary libraries
import seaborn as sns
import pandas as pd
import matplotlib.pyplot as plt
import matplotlib

# Load an example dataset
tips = sns.load_dataset("tips")
tips.head()

   total_bill   tip     sex smoker  day    time  size
0       16.99  1.01  Female     No  Sun  Dinner     2
1       10.34  1.66    Male     No  Sun  Dinner     3
2       21.01  3.50    Male     No  Sun  Dinner     3
3       23.68  3.31    Male     No  Sun  Dinner     2
4       24.59  3.61  Female     No  Sun  Dinner     4

2 Scatter Plot

A scatter plot is used to display the relationship between two continuous variables. Each point on the plot represents an observation in the dataset.

Code

sns.scatterplot(data=tips, x='tip', y='total_bill')

2.1 Color by Group

You can color the points in a scatter plot based on a categorical variable to visualize relationships within different groups.

Code

sns.scatterplot(data=tips, x='tip', y='total_bill', hue='sex')

2.2 Size by Group

Similarly, the size of the points can be varied based on a numerical or categorical variable.

Code

sns.scatterplot(data=tips, x='tip', y='total_bill', size='size')

3 Line Plot

A line plot is ideal for visualizing the trend of a continuous variable over a continuous interval or time.

Code

dowjones = sns.load_dataset("dowjones")
dowjones.head()

        Date  Price
0 1914-12-01  55.00
1 1915-01-01  56.55
2 1915-02-01  56.00
3 1915-03-01  58.30
4 1915-04-01  66.45

Code

sns.lineplot(data=dowjones, x='Date', y='Price')

3.1 Line Plot with Dots

You can also add markers to the line plot to highlight the data points. ::: {.cell}

Code

sns.lineplot(data=dowjones, x='Date', y='Price', marker='o')

:::

3.2 Color by Group

Different lines can be plotted for different categories to compare trends.

Code

import random

# Create datasets for comparison
dowjones2 = dowjones.copy()
dowjones2['type'] = 'old'

dowjones3 = dowjones.copy()
dowjones3['Price'] = dowjones3['Price'] + random.random() * 200
dowjones3['type'] = 'new'

dowjones4 = pd.concat([dowjones2, dowjones3], ignore_index=True)
dowjones4 = dowjones4.sort_values('Date').reset_index(drop=True)

Code

dowjones4.head()

        Date       Price type
0 1914-12-01   55.000000  old
1 1914-12-01  148.624455  new
2 1915-01-01  150.174455  new
3 1915-01-01   56.550000  old
4 1915-02-01   56.000000  old

Code

sns.lineplot(data=dowjones4, x='Date', y='Price', hue='type')

4 Histogram

A histogram is used to represent the distribution of a single numerical variable.

Code

sns.histplot(data=tips, x='tip')

4.1 Color by Group

Histograms can be grouped by a categorical variable to compare distributions.

Code

sns.histplot(data=tips, x='tip', hue='sex', multiple="dodge")

5 Bar Chart

A bar chart represents categorical data with rectangular bars. The lengths of the bars are proportional to the values they represent.

Code

sns.barplot(data=tips, x='sex', y='tip', errorbar=None)

5.1 Show Number on Bars

You can display the value of each bar directly on the plot.

Code

ax = sns.barplot(data=tips, x='sex', y='tip', errorbar=None)

for i in ax.containers:
    ax.bar_label(i,)

5.2 Horizontal Bar Plot

Bar charts can also be plotted horizontally.

Code

ax = sns.barplot(data=tips, y='sex', x='tip', errorbar=None, orient='h')
plt.show()

6 Box Plot

A box plot displays the five-number summary of a set of data: minimum, first quartile, median, third quartile, and maximum.

Code

sns.boxplot(data=tips, x='day', y='tip')

6.1 Color by Group

Box plots can be grouped by a categorical variable to compare the distributions.

Code

sns.boxplot(data=tips, x='day', y='tip', hue='sex')

7 Strip Plot

A strip plot is a scatter plot where one of the variables is categorical. It is useful for visualizing the distribution of data points.

Code

sns.stripplot(data=tips, x='day', y='tip')

7.1 Color by Group

Strip plots can also be grouped by a categorical variable.

Code

sns.stripplot(data=tips, x='day', y='tip', hue='sex', dodge=True)

8 Joint Plot

A joint plot shows the relationship between two variables along with their individual distributions.

Code

sns.jointplot(data=tips, x='total_bill', y='tip', kind='reg')

<seaborn.axisgrid.JointGrid object at 0x1118ed110>

9 Facet Plot

Facet plots allow you to create multiple plots based on the subsets of your data.

Code

g = sns.FacetGrid(data=tips, col="day", hue="sex")
g.map_dataframe(sns.scatterplot, x="total_bill", y="tip")

Code

g.add_legend()

9.1 Two Plots per Column

You can wrap the columns of the facet grid to control the layout.

Code

g = sns.FacetGrid(data=tips, col="day", col_wrap=2, hue="sex")
g.map_dataframe(sns.scatterplot, x="total_bill", y="tip")

Code

g.add_legend()

10 Subplots

You can create a figure with multiple subplots to display several plots at once.

Code

fig, axes = plt.subplots(1, 2)

sns.boxplot(data=tips, x='day', y='tip', hue='sex', ax=axes[0])
sns.boxplot(data=tips, x='day', y='tip', ax=axes[1])

11 Displaying Chinese Characters on macOS

To display Chinese characters correctly in plots on macOS, you need to set the font family to one that supports them.

Code

# Add the following line
plt.rcParams['font.family'] = ['Arial Unicode MS'] # To display Chinese labels correctly
plt.rcParams['axes.unicode_minus'] = False # To display the minus sign correctly
sns.set_style('whitegrid', {'font.sans-serif': ['Arial Unicode MS', 'Arial']})

12 Customizing Title, Size, and Axis Labels

12.1 Add Title

You can add a title to your plot to describe its content.

Code

df = sns.load_dataset("tips")
ax = sns.boxplot(x="day", y="total_bill", data=df)
ax.set_title("Tips Box Plot")

12.2 Adjust Size

The size of the plot can be adjusted to fit your needs.

Code

plt.clf()
plt.figure(figsize=(10, 6))
ax = sns.boxplot(x="day", y="total_bill", data=df)
ax.set_title("Tips Box Plot")
plt.show()

12.3 Change Axis Labels

You can customize the labels for the x and y axes.

Code

ax = sns.boxplot(x="day", y="total_bill", data=df)
ax.set_title("Tips Box Plot")
ax.set(xlabel='X-axis Label', ylabel='Y-axis Label')

13 Applying Themes

Seaborn comes with several built-in themes to style your plots.

The “darkgrid” theme is the default and features a gray background with white grid lines.

Code

import seaborn as sns
df = sns.load_dataset("tips")
sns.set_theme()
# Equivalent to:
# sns.set_style("darkgrid")
sns.boxplot(x="day", y="total_bill", data=df)

The “whitegrid” theme has a white background with gray grid lines.

Code

import seaborn as sns
df = sns.load_dataset("tips")
sns.set_style("whitegrid")
sns.boxplot(x="day", y="total_bill", data=df)

The “dark” theme is similar to “darkgrid” but without the grid lines.

Code

import seaborn as sns
df = sns.load_dataset("tips")
sns.set_style("dark")
sns.boxplot(x="day", y="total_bill", data=df)

The “white” theme is similar to “whitegrid” but without the grid lines.

Code

import seaborn as sns
df = sns.load_dataset("tips")
sns.set_style("white")
sns.boxplot(x="day", y="total_bill", data=df)

The “ticks” theme is like the “white” theme but adds ticks to the axes.

Code

import seaborn as sns
df = sns.load_dataset("tips")
sns.set_style("ticks")
sns.boxplot(x="day", y="total_bill", data=df)

This theme mimics the style of the FiveThirtyEight website.

Code

plt.clf()
plt.style.use('fivethirtyeight')
sns.boxplot(x="day", y="total_bill", data=df)
plt.show()

This theme emulates the popular ggplot2 library in R.

Code

plt.clf()
plt.style.use('ggplot')
sns.boxplot(x="day", y="total_bill", data=df)
plt.show()

This theme uses a color palette that is friendly to colorblind viewers.

Code

plt.clf()
plt.style.use('tableau-colorblind10')
sns.boxplot(x="day", y="total_bill", data=df)
plt.show()

This theme uses a dark background for the plots.

Code

plt.clf()
plt.style.use('dark_background')
sns.boxplot(x="day", y="total_bill", data=df)
plt.show()

14 Saving a Plot

You can save your plot to a file in various formats.

Code

import seaborn as sns
df = sns.load_dataset("tips")
plt.clf()
plt.style.use('default')
sns.boxplot(x="day", y="total_bill", data=df)
plt.savefig("output.png", dpi=100, bbox_inches="tight")

15 Animated Plot

You can create animated plots to show changes over time or another variable.

Code

from celluloid import Camera

Code

from celluloid import Camera
from matplotlib import pyplot as plt

fig = plt.figure()
camera = Camera(fig)

a = sns.lineplot(data=dowjones4, x='Date', y='Price', hue='type')
hands, labs = a.get_legend_handles_labels()

new_data = dowjones4.sample(50, random_state=42)
new_data = new_data.sort_values(by=['Date'], ascending=True)

for i in new_data["Date"]:
    data = dowjones4.query('Date <= @i')
    sns.lineplot(data=data, x='Date', y='Price', hue='type')
    plt.legend(handles=hands, labels=labs)
    camera.snap()

animation = camera.animate()

Code

from IPython.display import HTML
HTML(animation.to_html5_video())

16 References

--- title: "Seaborn Chart" execute: warning: false error: false format: html: toc: true toc-location: right code-fold: show code-tools: true number-sections: true code-block-bg: true code-block-border-left: "#31BAE9" --- ![](images/logo-wide-lightbg.svg) ```{r} library(reticulate) py_require(c('celluloid','seaborn','IPython')) ``` # Introduction to Seaborn Seaborn is a Python data visualization library based on Matplotlib. It provides a high-level interface for creating attractive and informative statistical graphics. This document demonstrates how to create various plots using Seaborn, customize their appearance, and save them. ```{python} import seaborn as sns print(sns.__version__) ``` ```{python} # Import necessary libraries import seaborn as sns import pandas as pd import matplotlib.pyplot as plt import matplotlib # Load an example dataset tips = sns.load_dataset("tips") tips.head() ``` # Scatter Plot A scatter plot is used to display the relationship between two continuous variables. Each point on the plot represents an observation in the dataset. ```{python} sns.scatterplot(data=tips, x='tip', y='total_bill') ``` ## Color by Group You can color the points in a scatter plot based on a categorical variable to visualize relationships within different groups. ```{python} sns.scatterplot(data=tips, x='tip', y='total_bill', hue='sex') ``` ## Size by Group Similarly, the size of the points can be varied based on a numerical or categorical variable. ```{python} sns.scatterplot(data=tips, x='tip', y='total_bill', size='size') ``` # Line Plot A line plot is ideal for visualizing the trend of a continuous variable over a continuous interval or time. ```{python} dowjones = sns.load_dataset("dowjones") dowjones.head() ``` ```{python} sns.lineplot(data=dowjones, x='Date', y='Price') ``` ## Line Plot with Dots You can also add markers to the line plot to highlight the data points. ```{python} sns.lineplot(data=dowjones, x='Date', y='Price', marker='o') ``` ## Color by Group Different lines can be plotted for different categories to compare trends. ```{python} #| code-fold: true import random # Create datasets for comparison dowjones2 = dowjones.copy() dowjones2['type'] = 'old' dowjones3 = dowjones.copy() dowjones3['Price'] = dowjones3['Price'] + random.random() * 200 dowjones3['type'] = 'new' dowjones4 = pd.concat([dowjones2, dowjones3], ignore_index=True) dowjones4 = dowjones4.sort_values('Date').reset_index(drop=True) ``` ```{python} dowjones4.head() ``` ```{python} sns.lineplot(data=dowjones4, x='Date', y='Price', hue='type') ``` # Histogram A histogram is used to represent the distribution of a single numerical variable. ```{python} sns.histplot(data=tips, x='tip') ``` ## Color by Group Histograms can be grouped by a categorical variable to compare distributions. ```{python} sns.histplot(data=tips, x='tip', hue='sex', multiple="dodge") ``` # Bar Chart A bar chart represents categorical data with rectangular bars. The lengths of the bars are proportional to the values they represent. ```{python} sns.barplot(data=tips, x='sex', y='tip', errorbar=None) ``` ## Show Number on Bars You can display the value of each bar directly on the plot. ```{python} ax = sns.barplot(data=tips, x='sex', y='tip', errorbar=None) for i in ax.containers: ax.bar_label(i,) ``` ## Horizontal Bar Plot Bar charts can also be plotted horizontally. ```{python} ax = sns.barplot(data=tips, y='sex', x='tip', errorbar=None, orient='h') plt.show() ``` # Box Plot A box plot displays the five-number summary of a set of data: minimum, first quartile, median, third quartile, and maximum. ```{python} sns.boxplot(data=tips, x='day', y='tip') ``` ## Color by Group Box plots can be grouped by a categorical variable to compare the distributions. ```{python} sns.boxplot(data=tips, x='day', y='tip', hue='sex') ``` # Strip Plot A strip plot is a scatter plot where one of the variables is categorical. It is useful for visualizing the distribution of data points. ```{python} sns.stripplot(data=tips, x='day', y='tip') ``` ## Color by Group Strip plots can also be grouped by a categorical variable. ```{python} sns.stripplot(data=tips, x='day', y='tip', hue='sex', dodge=True) ``` # Joint Plot A joint plot shows the relationship between two variables along with their individual distributions. ```{python} sns.jointplot(data=tips, x='total_bill', y='tip', kind='reg') ``` # Facet Plot Facet plots allow you to create multiple plots based on the subsets of your data. ```{python} g = sns.FacetGrid(data=tips, col="day", hue="sex") g.map_dataframe(sns.scatterplot, x="total_bill", y="tip") g.add_legend() ``` ## Two Plots per Column You can wrap the columns of the facet grid to control the layout. ```{python} g = sns.FacetGrid(data=tips, col="day", col_wrap=2, hue="sex") g.map_dataframe(sns.scatterplot, x="total_bill", y="tip") g.add_legend() ``` # Subplots You can create a figure with multiple subplots to display several plots at once. ```{python} fig, axes = plt.subplots(1, 2) sns.boxplot(data=tips, x='day', y='tip', hue='sex', ax=axes[0]) sns.boxplot(data=tips, x='day', y='tip', ax=axes[1]) ``` # Displaying Chinese Characters on macOS To display Chinese characters correctly in plots on macOS, you need to set the font family to one that supports them. ```{python} # Add the following line plt.rcParams['font.family'] = ['Arial Unicode MS'] # To display Chinese labels correctly plt.rcParams['axes.unicode_minus'] = False # To display the minus sign correctly sns.set_style('whitegrid', {'font.sans-serif': ['Arial Unicode MS', 'Arial']}) ``` # Customizing Title, Size, and Axis Labels ## Add Title You can add a title to your plot to describe its content. ```{python} df = sns.load_dataset("tips") ax = sns.boxplot(x="day", y="total_bill", data=df) ax.set_title("Tips Box Plot") ``` ## Adjust Size The size of the plot can be adjusted to fit your needs. ```{python} plt.clf() plt.figure(figsize=(10, 6)) ax = sns.boxplot(x="day", y="total_bill", data=df) ax.set_title("Tips Box Plot") plt.show() ``` ## Change Axis Labels You can customize the labels for the x and y axes. ```{python} ax = sns.boxplot(x="day", y="total_bill", data=df) ax.set_title("Tips Box Plot") ax.set(xlabel='X-axis Label', ylabel='Y-axis Label') ``` # Applying Themes Seaborn comes with several built-in themes to style your plots. ::: {.panel-tabset .nav-pills} ## Darkgrid Theme The "darkgrid" theme is the default and features a gray background with white grid lines. ```{python} import seaborn as sns df = sns.load_dataset("tips") sns.set_theme() # Equivalent to: # sns.set_style("darkgrid") sns.boxplot(x="day", y="total_bill", data=df) ``` ## Whitegrid Theme The "whitegrid" theme has a white background with gray grid lines. ```{python} import seaborn as sns df = sns.load_dataset("tips") sns.set_style("whitegrid") sns.boxplot(x="day", y="total_bill", data=df) ``` ## Dark Theme The "dark" theme is similar to "darkgrid" but without the grid lines. ```{python} import seaborn as sns df = sns.load_dataset("tips") sns.set_style("dark") sns.boxplot(x="day", y="total_bill", data=df) ``` ## White Theme The "white" theme is similar to "whitegrid" but without the grid lines. ```{python} import seaborn as sns df = sns.load_dataset("tips") sns.set_style("white") sns.boxplot(x="day", y="total_bill", data=df) ``` ## Ticks Theme The "ticks" theme is like the "white" theme but adds ticks to the axes. ```{python} import seaborn as sns df = sns.load_dataset("tips") sns.set_style("ticks") sns.boxplot(x="day", y="total_bill", data=df) ``` ## fivethirtyeight Theme This theme mimics the style of the FiveThirtyEight website. ```{python} plt.clf() plt.style.use('fivethirtyeight') sns.boxplot(x="day", y="total_bill", data=df) plt.show() ``` ## ggplot Theme This theme emulates the popular `ggplot2` library in R. ```{python} plt.clf() plt.style.use('ggplot') sns.boxplot(x="day", y="total_bill", data=df) plt.show() ``` ## tableau-colorblind10 Theme This theme uses a color palette that is friendly to colorblind viewers. ```{python} plt.clf() plt.style.use('tableau-colorblind10') sns.boxplot(x="day", y="total_bill", data=df) plt.show() ``` ## dark_background Theme This theme uses a dark background for the plots. ```{python} plt.clf() plt.style.use('dark_background') sns.boxplot(x="day", y="total_bill", data=df) plt.show() ``` ::: # Saving a Plot You can save your plot to a file in various formats. ```{python} import seaborn as sns df = sns.load_dataset("tips") plt.clf() plt.style.use('default') sns.boxplot(x="day", y="total_bill", data=df) plt.savefig("output.png", dpi=100, bbox_inches="tight") ``` # Animated Plot You can create animated plots to show changes over time or another variable. ```{python} from celluloid import Camera ``` ```{python} #| output: false from celluloid import Camera from matplotlib import pyplot as plt fig = plt.figure() camera = Camera(fig) a = sns.lineplot(data=dowjones4, x='Date', y='Price', hue='type') hands, labs = a.get_legend_handles_labels() new_data = dowjones4.sample(50, random_state=42) new_data = new_data.sort_values(by=['Date'], ascending=True) for i in new_data["Date"]: data = dowjones4.query('Date <= @i') sns.lineplot(data=data, x='Date', y='Price', hue='type') plt.legend(handles=hands, labels=labs) camera.snap() animation = camera.animate() ``` ```{python} from IPython.display import HTML HTML(animation.to_html5_video()) ``` # References - [Seaborn Official Documentation](https://seaborn.pydata.org/index.html) - [YouTube Tutorial on Seaborn](https://www.youtube.com/watch?v=ooqXQ37XHMM)