This section is meant to help users on well-supported platforms and common Python environments get started using Chaco as quickly as possible. Chaco users can subscribe to the enthought-dev mailing list to post questions, consult archives, and share tips.
There are several ways to get Chaco. The easiest way is through the Enthought Python Distribution (EPD), which is available for several platforms and also provides many other useful packages. Chaco may also be available through a package manager on your platform, such as apt on Ubuntu or MacPorts on OS X. You can also build Chaco yourself, but because of the number of packages required, we highly recommend you install EPD.
Chaco, the rest of the Enthought Tool Suite, and a lot more are bundled with EPD. Getting EPD allows you to install Chaco and all its dependencies at once; however, these packages will be linked to a new instance of Python. The EPD Free distribution is free for all users and contains all that you need to use Chaco.
To get EPD, go to the EPD download page and get the appropriate version for your platform. After running the installer, you will have a working version of Chaco and several examples.
Building Chaco on your machine requires you to build Chaco and each of its dependencies, but it has the advantage of installing Chaco on top of the Python instance you already have installed. The build process may be challenging and will require you to have SWIG, Cython and several development libraries installed.
To do this, you can either
Chaco ships with several examples for testing your installation and to show you what Chaco can do. Almost all of the examples are stand-alone files that you can run individually, from any location. Depending on how you installed Chaco, you may or may not have the examples already.
If you installed Chaco as part of EPD, the location of the examples depends on your platform:
If you downloaded and installed Chaco from source (from GitHub or via the PyPI tar.gz file), the examples are located in the examples/ subdirectory inside the root of the Chaco source tree, next to docs/ and the enthought/ directories.
If you don’t know how Chaco was installed, you can download the latest versions of examples individually from github:
For ETS 3.0 or Chaco 3.0, you can check out the examples with Subversion:
svn co https://svn.enthought.com/svn/enthought/Chaco/tags/3.0.0/examples
For ETS 2.8 or Chaco 2.0.x:
svn co https://svn.enthought.com/svn/enthought/Chaco/tags/enthought.chaco2_2.0.5/examples
Chaco examples can be found in the examples/demo/ and examples/tutorials/ directories. Some are classified by themes and located in separate directories. Almost all of the Chaco examples are standalone files that can be run individually. We will first show how to execute them from the command line, and then we will show how to run Chaco in an interactive way from IPython. This “shell” mode will be more familiar to Matplotlib or Matlab users.
Some of these examples can be visualized in our Chaco gallery.
From the examples/demo directory, run the simple_line example:
This opens a plot of several Bessel functions with a legend.
You can interact with the plot in several ways: .. Ctrl-Left and Ctrl-Right don’t work in OS X?
To pan the plot, hold down the left mouse button inside the plot area (but not on the legend) and drag the mouse.
To zoom the plot:
- Mouse wheel: scroll up to zoom in, and scroll down to zoom out (or the reverse you’re on a version of OS X with ‘natural scrolling’).
- Zoom box: Press z, and then draw a box region to zoom in on. (There is no box-based zoom out.) Press Ctrl-Left and Ctrl-Right to go back and forward in your zoom box history.
- Drag: hold down the right mouse button and drag the mouse up or down. Up zooms in, and down zooms out.
- For any of the above, press Escape to reset the zoom to the original view.
To move the legend, hold down the right mouse button inside the legend and drag it around. Note that you can move the legend outside of the plot area.
To exit the plot, click the “close window” button on the window frame or (on Mac) choose the Quit option on the Python menu. Alternatively, can you press Ctrl-C in the terminal.
You can run most of the examples in the the examples/demo/basic/ directory and the examples/demo/shell/ directory. The examples/demo/advanced/ directory has some examples that require additional data or packages. In particular,
For detailed information about each built-in example, see the Annotated Examples section.
While all of the Chaco examples can be launched from the command line using the standard Python interpreter, if you have IPython installed, you can poke around them in a more interactive fashion.
Chaco provides a subpackage, currently named the “Chaco Shell”, for doing command-line plotting like Matlab or Matplotlib. The examples in the examples/demo/shell/ directory use this subpackage, and they are particularly amenable to exploration with IPython.
The first example we’ll look at is the lines.py example. First, we’ll run it using the standard Python interpreter:
This shows two overlapping line plots.
You can interact with this plot just as in the previous section.
Now exit the plot, and start IPython with the --gui=wx option :
This tells IPython to start a wxPython mainloop in a background thread. Now run the previous example again:
In : run lines.py
This displays the plot window, but gives you another IPython prompt. You can now use various commands from the chaco.shell package to interact with the plot.
Import the shell commands:
In : from chaco.shell import *
Set the X-axis title:
In : xtitle("X data")
Toggle the legend:
In : legend()
After running these commands, your plot looks like this:
The chaco_commands() function display a list of commands with brief descriptions.
You can explore the Chaco object hierarchy, as well. The chaco.shell commands are just convenience functions that wrap a rich object hierarchy that comprise the actual plot. See the Exploring Chaco with IPython section for information on all you can do with Chaco from within IPython.
Let’s create, from scratch, the simplest possible Chaco plot (embedded inside a Traits application).
First, some imports to bring in necessary components:
from chaco.api import ArrayPlotData, Plot from enable.component_editor import ComponentEditor from traits.api import HasTraits, Instance from traitsui.api import View, Item
The imports from chaco and enable support the creation of the plot. The imports from traits bring in components to embed the plot inside a Traits application. (Refer to the Traits documentation for more details about building an interactive application using Traits.) Now let’s create a Traits class with a view that contains only one element: a Chaco plot inside a slightly customized window:
class MyPlot(HasTraits): plot = Instance(Plot) traits_view = View(Item('plot', editor = ComponentEditor(), show_label = False), width = 500, height = 500, resizable = True, title = "My line plot")
A few options have been set to control the window containing the plot. Now, when the plot is created, we would like to pass in our data. Let’s assume the data is a set of points with coordinates contained in two NumPy arrays x and y. So, adding an __init__ method to create the Plot object looks as follows:
class MyPlot(HasTraits): plot = Instance(Plot) traits_view = View(Item('plot', editor = ComponentEditor(), show_label = False), width = 500, height = 500, resizable = True, title = "My line plot") def __init__(self, x, y, *args, **kw): super(MyPlot, self).__init__(*args, **kw) plotdata = ArrayPlotData(x=x,y=y) plot = Plot(plotdata) plot.plot(("x","y"), type = "line", color = "blue") plot.title = "sin(x)*x**3" self.plot = plot
Since it inherits from HasTraits, the new class can use all the power of Traits, and the call to super() in its __init__ method makes sure this object possesses the attributes and methods of its parent class. Now let’s use our Traits object. Below, we generate some data, pass it to an instance of MyPlot and call configure_traits to create the UI:
import numpy as np x = np.linspace(-14,14,100) y = np.sin(x)*x**3 lineplot = MyPlot(x,y) lineplot.configure_traits()
The result should look like
This might look like a lot of code to visualize a function, but this is a relatively simple basis on top of which we can build full-featured applications with custom UIs and custom tools. For example, the Traits object allows you to create controls for your plot at a very high level, add these controls to the UI with very little work, and add listeners to update the plot when the data changes. Chaco also allows you to create tools to interact with the plot and overlays that make these tools intuitive and visually appealing.
|||Starting from IPython 0.12, it is possible to use the Qt backend with --gui=qt. Make sure that the environment variable QT_API is set correctly, as described here|