Programming simple CA interactively

zasim is, at heart, a package of useful utilities and parts and pieces to experiment with cellular automata. For experimentation, an interactive console is a good tool to find out what zasim offers and to experiment with different step functions.

There is additional support for running zasim with the new IPython version 0.11 or higher, which offers a “rich” console and a web application called “notebook”. Using the ipython qtconsole, images from the cellular automaton can be displayed directly inside the console as pictures, rather than ascii art.

The package zasim.cagen offers a variety of pre-fabricated simulator prototypes with interesting parameters you can tweak and the zasim.display package offers a way to display each step of the simulator on the console as ascii art - or as images when using the rich consoles from IPython.

Starting out: 1d CA

Starting out with zasim is quite easy. First, open up your favorite interactive console, be it the vanilla python interpreter, bpython, IPython or anything else. You can then go on and import the cagen package from zasim and the console module from the display package. In order to create our own starting configuration, we also need numpy, which is usually imported under the name “np”.

>>> from zasim import cagen
>>> from zasim.display.console import OneDimConsolePainter
>>> import numpy as np

The most interesting classes for now are zasim.cagen.simulators.ElementarySimulator and zasim.display.console.OneDimConsolePainter. The first one is simply a class for one- or twodimensional elementary cellular automata - in this case this just refers to cellular automata that are identified by the used neighbourhood and a rule number - and the second one is a class that displays the contents of the simulators configuration space as ascii art.

In order to set up a simulator, all we need is to instantiate an ElementarySimulator object. It will take at least the rule number and either a size to be used for a random configuration or the configuration as a numpy array. In this case we will create a known configuration. In order to display it, too, we can create a OneDimConsolePainter and pass the simulator as the first argument. It will then, by default, automatically print out every step that happens in the simulator on the console.

We will just go ahead and choose rule 126, which will paint a sierpinski triangle. As starting configuration we use a single one surrounded by lots of zeros. The code looks like this:

>>> config = np.array([0] * 30 + [1] + [0] * 30)
>>> sim = cagen.ElementarySimulator(config=config, rule=126)
>>> disp = OneDimConsolePainter(sim, lines=1)
                              #
>>> sim.step()
                             ###
>>> sim.step()
                            ## ##

As you can see, the config gets printed once when the display object is created and then after each step. The same goes for stepping in a loop:

>>> for i in range(10): sim.step()
...
                           #######
                          ##     ##
                         ####   ####
                        ##  ## ##  ##
                       ###############
                      ##             ##
                     ####           ####
                    ##  ##         ##  ##
                   ########       ########
                  ##      ##     ##      ##

Now to explain the lines, one by one:

>>> config = np.array([0] * 30 + [1] + [0] * 30)

This simply creates a numpy array from thirty zeros, one one and another 30 zeros. This will give us a prettier picture than the random configuration we would have gotten, had we supplied the size argument, rather than a configuration.

>>> sim = cagen.ElementarySimulator(config=config, rule=126)

The ElementarySimulator takes as arguments the configuration to use or a size, if the config should be randomly created, as well as a rule number and then some extra options that are not interesting to us right now. Those are almost the same as the one you can supply on the commandline to the zasim.cagen.main module.

In this case we create such a simulator from the config we built and set the rule number to use to 126.

>>> disp = OneDimConsolePainter(sim, lines=1)

The OneDimConsolePainter takes as first argument the simulator to take configurations from and the lines keyword argument controls how many lines are to be stored in the display. For interactive console use, 1 is a good value, because otherwise, after each step, the OneDimConsolePainter would print out its complete data.

Each call to sim.step will afterwards run the step function on the configuration and signal all connected displays - in this case just the display. Since we supplied the default value for OneDimConsolePainter while constructing it, it has connect and auto_output set to true. The connect parameter tells the display to directly connect to the changed and updated signals of the simulator. The other tells the display to output its data after every change.

IPython interactivity helpers

If you use the new IPython with qtconsole or the notebook web-app, you can also display configurations in-line, right inside the applications, as images or HTML. This is how that works:

For a html-based display, you can just set auto_output for the OneDimConsolePainter to false and display the configuration like this:

>>> disp
[the configuration would be displayed here]

This does not work in functions or loops, however, because it relies on the interpreter automatically displaying the representation of any value that is not caught. If you want to force the display, you can just import the display function from IPython directtly:

>>> from IPython.core.display import display
>>> display(disp)
[the configuration would be displayed here]

If you want to show the config as a picture, rather than an HTML table, you can use the OneDimQImagePainter instead, which works much like the OneDimConsolePainter:

>>> from zasim.display.qt import OneDimQImagePainter
>>> disp = OneDimQImagePainter(sim, lines=50, scale=4)
>>> disp
[the configuration would be displayed here]

For a QImage based painter, it is much more sensible to use a higher lines value, because this way, the image would be a bit bigger. Note though, that the position the configuration is painted to travels downwards and is wrapped from the bottom back up to the top, so sometimes you will see the current configuration in the middle, older values above and even older values directly below.

The Game of Life - 2d CA

Although the ElementarySimulator supports 2d configurations as well, the Game of Life simulator is much nicer to look at in general. For our next adventure, we instantiate a GameOfLife and the matching TwoDimConsolePainter.

>>> from zasim import cagen
>>> from zasim.display.console import TwoDimConsolePainter
>>> import numpy as np

The most symbolic figure of Game of Life is probably the glider. We will create an empty configuration and paste a glider into it, as well as an obstacle for it to collide with:

>>> config = np.zeros((6, 11), dtype=int)
>>> config[0:3,1:4] = np.array([
...    [0,1,0],
...    [0,0,1],
...    [1,1,1]])
>>> config[3:6, 9] = [1, 1, 1]
>>> config
array([[0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0],
       [0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0],
       [0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0],
       [0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0],
       [0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0],
       [0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0]])

And now we can put the configuration into the simulator, like this:

>>> sim = cagen.GameOfLife(config=config)
>>> disp = TwoDimConsolePainter(sim)

  #
# #
 ##





   ###

Stepping a few times will show the typical glider movement, until it collides with the blocks at the end. This specific configuration will eventually die out.

Table Of Contents

Previous topic

Invoking zasim from the commandline

Next topic

Assembling a custom StepFunc

This Page