What is Python Bokeh?

Key takeaways:

Bokeh simplifies interactive data visualization in Python for web browsers, providing both flexibility and scalability.
Bokeh offers powerful modules for creating customized visualizations, such as line plots, scatter plots, and more.
Bokeh’s features include exporting visuals in formats like PNG and PDF, as well as interactive capabilities like panning and zooming.
Bokeh integrates seamlessly with popular Python libraries like pandas and NumPy, enhancing data analysis workflows.

Bokeh for data visualization

Bokeh is a Python library for creating interactive visualizations. It provides powerful tools for exploring various data insights, offering flexibility, interactivity, and scalability.

Step 3: How to perform data visualization with Bokeh

Bokeh is a vast library that offers a variety of plots, graphs, and charts to create a variety of visuals. Different visual diagrams can be created to represent the statistical information of datasets. Following are a few from the wide range of statistical visuals created through Bokeh:

Want to learn and explore the concepts of data visualization in a real-world application? Try out the Stock Market Data Visualization Using Python project.

Code example

Here is a code example that exhibits the range of visualization tools offered by Bokeh.

# Import Libraries
import numpy as np
import pandas as pd
from math import pi
from bokeh.io import output_file, save
from bokeh.plotting import figure, show
from bokeh.layouts import gridplot
from bokeh.transform import cumsum

# Specify range
rng = np.random.default_rng()
sampleData = rng.normal(loc=0, scale=1, size=500)

# 2. Histogram
histPlot = figure(width=670, height=400, toolbar_location=None, title="Impact of spice level on CRC")

bins = np.linspace(-4, 4, 30)
hist, edges = np.histogram(sampleData, density=True, bins=bins)
histPlot.quad(top=hist, bottom=0, left=edges[:-1], right=edges[1:], fill_color="purple", line_color="white", legend_label="500 random samples")

# Labels
histPlot.y_range.start = 0
histPlot.xaxis.axis_label = "Spice level"
histPlot.yaxis.axis_label = "CRC"

# 2. Line plot
linePlot = figure(width=670, height=400, toolbar_location=None, title="Line Plot")
x = np.arange(0, 10, 0.1)
y = np.sin(x)
linePlot.line(x, y, line_width=2, color="green", legend_label="Sine Wave")

# 3. Scatter plot
scatterPlot = figure(width=670, height=400, toolbar_location=None, title="Scatter Plot")
scatter_x = rng.normal(size=100)
scatter_y = rng.normal(size=100)
scatterPlot.scatter(scatter_x, scatter_y, size=10, color="blue", alpha=0.5, legend_label="Random Scatter")

# 4. Box plot
boxPlot = figure(width=670, height=400, toolbar_location=None, title="Box Plot")
boxPlot.vbar(x=[1, 2, 3], width=0.5, bottom=0, top=[np.median(sampleData)] * 3, color="orange", legend_label="Median")

# 5. Pie chart
data = pd.Series([10, 20, 30, 40], index=['Category A', 'Category B', 'Category C', 'Category D'])
data = data.reset_index(name='value').rename(columns={'index': 'category'})
data['angle'] = data['value'] / data['value'].sum() * 2 * pi
data['color'] = ['red', 'green', 'blue', 'orange']

piePlot = figure(width=400, height=400, toolbar_location=None, title="Pie Chart")
piePlot.wedge(x=0, y=1, radius=0.4, start_angle=cumsum('angle', include_zero=True), end_angle=cumsum('angle'),
              line_color="white", fill_color='color', legend_field='category', source=data)

# 6. Donut chart
donut_data = pd.Series([10, 20, 30, 40], index=['Category A', 'Category B', 'Category C', 'Category D'])
donut_data = donut_data.reset_index(name='value').rename(columns={'index': 'category'})
donut_data['angle'] = donut_data['value'] / donut_data['value'].sum() * 2 * pi
donut_data['color'] = ['red', 'green', 'blue', 'orange']

donutPlot = figure(width=400, height=400, toolbar_location=None, title="Donut Chart")
donutPlot.annular_wedge(x=0, y=1, inner_radius=0.2, outer_radius=0.4,
                        start_angle=cumsum('angle', include_zero=True),
                        end_angle=cumsum('angle'), line_color="white",
                        fill_color='color', legend_field='category', source=donut_data)

# 7. Combine all in a grid plot
grid = gridplot([[histPlot, linePlot], [scatterPlot, boxPlot], [piePlot, donutPlot]])

# Save and show plots
output_file("output.html")
show(grid)

An illustration of different charts offered by Bokeh Python

The explanation of the code is as follows:

Lines 2–8: Import essential libraries (numpy, pandas) for numerical operations, random number generation, and Bokeh modules for creating and displaying various types of plots.
Lines 10–12: Generate 500 random samples from a normal distribution using numpy’s random number generator (rng).
Lines 14–24 (Histogram Plot): Create a histogram figure with bins calculated using numpy.linspace. Compute the histogram values and edges, then plot using quad with purple bars. Set the x-axis as “Spice level” and the y-axis as “CRC,” ensuring the y-axis starts from 0.
Lines 26–30 (Line Plot): Create a sine wave plot by generating x values from 0 to 10 at intervals of 0.1. Compute y as the sine of x and plot it as a green line with a legend.
Lines 32–36 (Scatter Plot): Generate random x and y values for 100 points. Plot them as a scatter plot with blue markers and 50% transparency, adding a legend.
Lines 38–40 (Box Plot): Create a box plot with vertical bars (vbar) for three categories. The height of each bar is set to the median of the random sample data.
Lines 42–50 (Pie Chart): Prepare data for a pie chart with categories and values. Add columns for angles (proportional to values) and colors. Plot the pie chart using wedge, specifying start and end angles for each slice.
Lines 52–62 (Donut Chart): Similar to the pie chart, prepare data but include an inner radius to create a donut chart. Use annular_wedge to plot the chart with the same categories and colors as the pie chart.
Lines 64–65 (Grid Layout): Arrange all plots (histogram, line, scatter, box, pie, and donut charts) into a 3x2 grid layout using gridplot.
Lines 68–69: Save the plots to an HTML file and display the grid in a browser using show.

Key features

It allows exporting plots in different file formats such as PNG, SVG, and PDF.
It provides themes and styling options to customize the appearance of plots according to the required aesthetics.
It supports interactive panning and zooming on geographic plots to create interactive maps.
It offers a high-level API for the quick creation of a common plot with minimal code and a low-level API for more flexibility for advanced customization.
It generates JavaScript-based plots that can be displayed and easily shared in web browsers.

Key benefits of using Bokeh

Ease of use: Creates common plots, as well as custom and complex plots based on the requirements.
Shareable: Shares visual data easily that can also be rendered.
Integrable: Interacts with other popular Pydata tools very easily, e.g., with pandas and NumPy.
Interactive: Creates interactive plots that show a dynamic display based on the specified requirements, e.g., graph and color sliders.
Open source: Bokeh is an open-source project that the new developer can access.

Want to get hands-on practice with Bokeh? Try out the project Data Analysis and Visualization with sidetable and Bokeh, where we create visualizations and use them for data analysis.

Frequently asked questions

Haven’t found what you were looking for? Contact Us

What is bokeh in Python used for?

bokeh is used to create interactive and visually appealing and interactive data visualizations that can be displayed in web browsers.

What is the difference between matplotlib and bokeh?

Matplotlib creates static, simple plots, while bokeh generates interactive and web-based visualizations.

How to install Bokeh in Python?

You can install bokeh by running the command pip install bokeh in your terminal or command prompt.

What is the difference between Bokeh and Plotly?

Both create interactive plots, but bokeh focuses on high-performance, scalable visuals, while Plotly offers more built-in complex visualizations and dashboards.

Free Resources

Learn in-demand tech skills in half the time

PRODUCTS

Mock Interview

New

Courses

Skill Paths

Projects

Assessments