Glyphs in Bokeh

Bokeh is a library of Python which is used to create interactive data visualizations. In this article, we will discuss glyphs in Bokeh. But at first let’s see how to install Bokeh in Python.

Installation

To install this type the below command in the terminal.

conda install bokeh
Or
pip install bokeh  

Plotting with glyphs

Usually, a plot consists of geometric shapes either in the form of a line, circle, etc. So, Glyphs are nothing but visual shapes that are drawn to represent the data such as circles, squares, lines, rectangles, etc.  

Creating a basic line chart:

The line chart displays the visualization of x and y-axis points movements in the form of a line. To draw a line glyph to the figure, we use the line() method of the figure object.

Syntax:

 my_plot.line(a, b, line_width)  

Code: 

Python

# import the libraries
from bokeh.plotting import figure, show, output_file

# prepare some data
a = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
b = [2, 4, 6, 8, 10, 12, 14, 16, 18, 20]

# create a plot using figure
my_plot = figure(title="simple line chart", x_axis_label="X-Axis",
                 y_axis_label="Y-Axis")

# adding line graph
my_plot.line(a, b, line_width=3)

# display output in file
output_file("line.html")

# show the result
show(my_plot)

Output:

Combining multiple graphs

You can also add multiple graphs with the use of bokeh.plotting interface. To do so, you just need to call the line() function multiple times by passing different data as parameters as shown in the example.

Syntax:

p.line(x1, y2, legend_label, line_color, line_width)

Code:

Python

from bokeh.plotting import figure, show
from bokeh.io import output_notebook


# prepare some data
x1 = [1, 3, 4, 5, 6]
x2 = [5, 3, 8, 1, 8]
y1 = [6, 7, 8, 9, 4]
y2 = [3, 4, 5, 6, 7]

# create a new plot
p = figure(title="Drawing multiple lines",
           x_axis_label="X-Axis", y_axis_label="Y-Axis")

# add multiple renderers
p.line(x1, y1, legend_label="line 1", line_color="red", line_width=1)
p.line(x2, y2, legend_label="line 2", line_color="blue", line_width=1)
p.line(x1, y2, legend_label="line 3", line_color="black", line_width=1)

output_notebook()
show(p)

Output:

Rendering Circles

In order to add the circle glyph to your plot, we use the circle() method instead of the line() method used in the above example.

Circle(): We use this method to add a circle glyph to the plot. It takes x and y coordinates of the center as parameters. Apart from these, it takes parameters such as size, fill_color,fill_alpha,angle, line_color, line_alpha,radius,radius_dimensions,etc

Syntax:

p.circle(x, y, size, fill_color)

Criss_cross(): this method adds a circle glyph with a ‘+’ mark through the center of the circle and it takes x and y coordinates of the center.

Syntax:

p.circle_cross(x, y, size, fill_color, fill_alpha, line_width)

Circle_X(): This method adds a circle glyph with an ‘X’ mark through the center of the circle and it takes x and y coordinates of the center.

Syntax:

p.circle_x(x, y, size,fill_color, fill_alpha, line_width)

Code:

Python

from bokeh.plotting import figure, show
from bokeh.io import output_file

# prepare some data
x = [1, 2, 4, 6, 7]
y = [7, 6, 3, 9, 10]

# create a new plot with figure function
p = figure(title="Circle Glyph", plot_width=450, plot_height=400)

# create circle glyph
p.circle(x=x, y=y, size=25, fill_color="red")
p.circle_cross(x=[2, 4, 6, 8], y=[5, 8, 9, 11], size=25,
               fill_color="blue", fill_alpha=0.3, line_width=2)
p.circle_x(x=[4, 7, 2, 6], y=[7, 2, 4, 9], size=25,
           fill_color="green", fill_alpha=0.6, line_width=2)

# show the results
output_file('circle.html')
show(p)

Output:

Rendering Bars

Similarly, we can render bars using vbar() function for vertical bars and hbar() function for horizontal bars.

Creating vbar

To draw vbar, we specify center x-coordinate, bottom, and top endpoints as shown in the below example:

Syntax:

p.vbar(x, bottom, top,
color, width, fill_color,legend_label)

Code:   

Python

# Bokeh libraries
from bokeh.io import output_notebook
from bokeh.plotting import figure, show

# data
day = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
no_orders = [450, 628, 488, 210, 287,
             791, 508, 639, 397, 943]

# Output the visualization directly in the notebook
output_notebook()

# Create a figure
p = figure(title='Bar chart',
           plot_height=400, plot_width=600,
           x_axis_label='Day', y_axis_label='Orders Received',
           x_minor_ticks=2, y_range=(0, 1000),
           toolbar_location=None)

# The daily orders will be represented as vertical
# bars (columns)
p.vbar(x=day, bottom=0, top=no_orders,
       color='blue', width=0.75, fill_color='red',
       legend_label='Orders')

# Let's check it out
show(p)

Output:

Creating hbar

To draw hbar , we specify center y-coordinate, left and right endpoints, and height as shown in the below example:

Syntax:

p.hbar(y, height, left, right,
 color, width, fill_color,
legend_label)

Example Code:

Python

# Bokeh libraries
from bokeh.plotting import figure, show, output_file

# data
day = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
no_orders = [450, 628, 488, 210, 287, 791,
             508, 639, 397, 943]

# Create a figure
p = figure(title='Bar chart',
           plot_height=400, plot_width=600,
           x_axis_label='Orders Received', y_axis_label='Day',
           x_range=(0, 1000), toolbar_location=None)

# The daily orders will be represented as
# horizontal bars
p.hbar(y=day, height=0.5, left=0, right=no_orders,
       color='blue', width=0.75, fill_color='red',
       legend_label='Orders')

# Let's check it out
show(p)
output_file("ex.html")

Output:

Patch Glyph

The patch glyph shades a region of space in a particular color. We use the patch() method to develop a single patch and patches() method to develop multiple patches.

Single patch

Syntax:

p.patch(x, y, fill_color, line_color, alpha, line_width)

Code:

Python

# Bokeh libraries
from bokeh.plotting import figure, show, output_file

# data
x = [1, 2, 3, 4, 5]
y = [6, 7, 8, 5, 2]

# Create a figure
p = figure(title='Patch Glyph',
           plot_height=400, plot_width=600,
           x_axis_label='x', y_axis_label='y',
           toolbar_location=None)

# add patch glyph
p.patch(x, y, fill_color="blue", line_color='black',
        alpha=0.5, line_width=2)

# show the results
show(p)
output_file("ex.html")

Output:

Multiple Patches

Multiple patches can be created in a similar way by using the function patches() instead of patch(). We are passing the data in the form of a list of arrays for creating three patches with different colors in the example shown below.

Code:

Python

# Bokeh libraries
from bokeh.plotting import figure, show, output_file

# data
x = [[1, 2, 3], [4, 6, 8], [2, 4, 5, 4]]
y = [[2, 5, 6], [3, 6, 7], [2, 4, 7, 8]]

# Create a figure
p = figure(title='Patch Glyph',
           plot_height=400, plot_width=600,
           x_axis_label='x', y_axis_label='y',
           toolbar_location=None)

# add patches glyph
p.patches(x, y, fill_color=["blue", "green", "yellow"], line_width=2)

# show the results
show(p)
output_file("ex.html")

Output: