Python Hints

Contents

• Python Hints

  • Installation of required modules

  • References and tutorials

  • Notebooks

  • Sage

Python is a powerful object-oriented interpreted scripting/programming language. Some version of Python is almost certainly on your computer already (on unix, linux, OSX type:

$ python --version

to find out which version, or just:

$ python

to start a python shell with a prompt that looks like:

>>>

You may prefer to use IPython, which is a nicer shell than the pure python shell, with things like command completion and history. See the Quick IPython Tutorial.

Installation of required modules

To effectively use the pyclaw and Clawpack plotting routines that are written in Python, you will need version 2.7 or greater (but not 3.0 or above, which is not backwards compatible). some modules that are not included in the standard Python distribution. Python modules are loaded with the import statement in Python and a wide variety of specialized modules exist for various purposes since people use Python for many different purposes.

An alternative to installing the packages discussed below, you could also use the Clawpack Virtual Machine, which has all the required Python modules pre-installed.

For use with Clawpack, you will need the Numpy module (Numerical Python) that allows working with arrays in much the same way as in Matlab. This is distributed as part of SciPy (Scientific Python). See the Installing SciPy page for tips installing SciPy and NumPy on various platforms.

For plotting you will also need the matplotlib module which provides Matlab-like plotting commands for 1d and 2d plots (e.g. contour and pcolor plots). This is generally the hardest thing to get going properly. See the matplotlib Installation FAQ.

Often the easiest way to get all the modules you need is to install the

Anaconda Python Distribution or the Enthought Python Distribution, which is free for academic users. Versions are available for Windows, Mac OS X, and Redhat linux.

With some versions of Linux (e.g. debian and Ubuntu), you can easily install what’s needed using apt-get:

$ apt-get install python-numpy
$ apt-get install python-scipy
$ apt-get install python-matplotlib

For OS X, you might also try the Scipy Superpack. See also these tips on installing matplotlib on OS X.

References and tutorials

Some useful links to get started learning Python:

• Enthought Python Distribution

• Dive Into Python

• Python tutorial

• NumPy User Guide

• NumPy for Matlab users

• SciPy Reference Guide

• Matplotlib gallery

• `LeVeque’s class notes

  System Message: WARNING/2 (/var/folders/_s/dx0xgftn3_x04rdx0_w5nq7w0000gn/T/tmprm2dswmf/8ead17868f18cc8ee8df9726447f8b2fb9a2e536/python.rst, line 96); backlink

  Inline interpreted text or phrase reference start-string without end-string.

• <http://faculty.washington.edu/rjl/classes/am583s2013/notes/python.html>`_

• Langtangen’s book and Introductory slides

Notebooks

See IPython notebook examples.

Sage

Sage is an open source mathematics software collection with its own interface and notebook system. It is based on Python and the full distribution contains in particular the Python modules needed for Clawpack.

You can try Sage directly from the web without downloading, click on “Try Sage Online” link on the Sage webpage.

We are working on incorporating Pyclaw into Sage as an optional package.