[Tutor] Introduction to modelling with Python

[Eike Welk]

On Wednesday April 7 2010 11:38:07 AG wrote:

> 
> Eike
> 
> I just wanted to come back to you on the book recommendation you made
> "Python scripting for computational science" - I tracked down a cheapish
> copy of the 3rd edition from 2009 and flipping through it (it only
> arrived yesterday), it seems like it is going to be very useful.
> Certainly it draws a lot on numpy, goes into using Tcl for GUIs, and a
> number of recipes for scripting, regular expressions and so on ... lots
> to get my head around.  

"Python scripting for computational science" also contains a section about 
solving differential equations. It is in one of the chapters about additional 
libraries.

The book contains many Fortran examples. As an alternative to Fortran you 
should also look at Cython. This is a very fast, compiled companion language 
for Python. It produces *.so (*.dll) files that can be directly loaded into 
Python as modules. 
http://www.cython.org/

> With respect to my original question then,
> equipped with this book you recommended, a book on differential
> equations, and one on an intro to environmental modelling, that should
> give me enough to work on for the time being.

An other possible book for an inter library loan might be "Mathematical Models 
in Biology" from Leah Edelstein-Keshet. It contains many differential 
equations from different fields of biology. Additionally it contains a very 
good introduction how to write your own partial differential equations. 

It is not an ideal book for someone who wants to do numerical computations 
(like you), because the differential equations are solved symbolically with 
pen and paper. On the other hand: Why do numerical experiments, when you can 
find all possible solutions at once "just" with pen and paper. I believe 
however the real reason is, that the book was written before computers became 
cheap (1988).   

> 
> So, just wanted to close the circle by letting you know that I took your
> recommendation, and it looks like it will pay off in time.

Great!


Eike.