None


How to Create a Graph in Python

(1 vote, average 5.00 out of 5)

 

Python is an increasingly versatile language, and is frequently used for quickly writing up computation based programs.  It can be useful to graph input, and thankfully Python has one of the most convenient and accessible libraries to graph objects - Matplotlib.  Matplotlib is a plotting library for Python which extends the numpy library.  Using matplotlib we can quickly create a graph of any two lists.  It is a very powerful library capable of making many types of graphs, including (but not limited to) bar graphs, scatter plots, log graphs, and histograms. This tutorial will teach you how to very quickly make a graph in matplotlib.  Note that matplotlib is a very powerful library and this is just a scratch of the surface!

  1. First, you will need to install matplotlib if you do not already have it.  The matplotlib installation page provides a pretty comprehensive and easy walkthrough on installing matplotlib (and numpy if you do not already have it).  http://matplotlib.sourceforge.net/users/installing.html
  2. Now we're ready to write some code.  Begin by setting up a new figure in matplotlib, making sure to import the library. We'll also set up the "backend" of Matplotlib so we will be able to save the figure to our desired output.

 

1
2
3
import matplotlib.pyplot as plt
 
plt.figure()

  1. Calling the figure() method sets up a new figure for our current plot.  Depending on what kind of graph we want, we will call a different method.  Almost all graph types are called with two parameters for the x series and the y series, which are simply lists in Python.  Since most graphs are simple scatter plots, we will look at an example using plot().  In the example below, we declare each series statically.  Typically, one list will be some computation of the x_series.  To add multiple lines to a graph, simply call plot() again.  If you want to create a new graph, call figure() again.

 

4
5
6
7
8
9
10
11
#create some data
x_series = [0,1,2,3,4,5]
y_series_1 = [x**2 for x in x_series]
y_series_2 = [x**3 for x in x_series]
 
#plot the two lines
plt.plot(x_series, y_series_1)
plt.plot(x_series, y_series_2)

  1. Now we have plotted some data to our figure.  But how do we see it?  The most practical way is to save it to a .PDF or .PNG.
8
plt.savefig("example.png")

  1. Open the file you just saved to see your graph!

test

Adding Labels and Axis Limits

Of course, the graph above is accurate but it's not exactly useful.  Often times we want to want add labels and a title, change the axes, etc.  Here we will cover some of the most common things that can be added to a graph.  Let's re-visit the code from the first part and add some fun stuff.

 

1
2
3
4
5
6
7
8
9
10
11
#add in labels and title
plt.xlabel("Small X Interval")
plt.ylabel("Calculated Data")
plt.title("Our Fantastic Graph")
 
#add limits to the x and y axis
plt.xlim(0, 6)
plt.ylim(-5, 80)

#add legend
plt.legend()
 
#save figure
plt.savefig("example.pdf")

 

Most of these functions are pretty self-explanatory and won't be explained in any more depth here.  Now let's look at our graph:

example

 

Adding a Legend and Modifying Line Properties

When we are plotting multiple lines (or whatever type of graph you are plotting), it is often necessary to be able to distinguish the lines.  We can add labels to each line after we plot it, and then add a legend to the graph.  Here is an example:

 

1
2
3
4
plt.plot(x_series, y_series_1, label="x^2")
plt.plot(x_series, y_series_2, label="x^3")
 
plt.legend(loc="upper left")

example_copy

Putting It All Together

This tutorial has covered the very basics of how to create a graph in Python.  Browse The Tech Repo for more in-depth tutorials concerning Matplotlib.   Let's look at a final version of the code and the graph it creates.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
#set up matplotlib and the figure
import matplotlib.pyplot as plt
plt.figure
 
#create data
x_series = [0,1,2,3,4,5]
y_series_1 = [x**2 for x in x_series]
y_series_2 = [x**3 for x in x_series]
 
#plot data
plt.plot(x_series, y_series_1, label="x^2")
plt.plot(x_series, y_series_2, label="x^3")
 
#add in labels and title
plt.xlabel("Small X Interval")
plt.ylabel("Calculated Data")
plt.title("Our Fantastic Graph")
 
#add limits to the x and y axis
plt.xlim(0, 6)
plt.ylim(-5, 80)
 
#create legend
plt.legend(loc="upper left")
 
#save figure to png
plt.savefig("example.png")

example_copy_copy

 

Partner Links:
Last Updated on Monday, 19 July 2010 22:08  
Related Articles

» How to Install an Updated Driver For NVIDIA Graphics Cards

Whether you are putting a new graphics card into your computer or just want to update the driver, it is recommended that you visit the Manufacturers website. Here you will receive the latest update that will include many enhancements and improvements. Follow these steps to install the latest driver on your computer:Make sure your graphics card is installed or recognized by the system.Open up a web browser, like Internet Explorer, Mozilla Firefox, or Google Chrome, and visit...

» Conditional (If-Else) Statements in Python

Below is a quick example of how to use an if statement in python: 12345if x==3: print "We are inside an if statement" x = x + 1print "we are no longer in an if statement" Note that unlike most languages, no parentheses are needed, although including parentheses does not violate python syntax.  Anything indented after the if statement is included in the execution if the if statement evaluates to true.  The scope is closed once indentation returns to the indent location of the...

» How to Exit in Python

Exiting in python is fairly straight-forward.  You need to import the "sys" library, and then call exit from it.123import sys sys.exit()Calling exit with no arguments is a clean exit -- no error code is returned.  If you'd like to exit with an error message, simply pass the message as a parameter and exit will automatically exit with an error and print your message.123import sys sys.exit("This is an example of an error message")