Book Reviews


[Day Permalink] Wednesday, April 28, 2004

[Item Permalink] Learning Python for scientific computing -- Comment()
Today I wrote my first complete Python programs. The other one simulated placing pairs of points into the unit square, and was rather easy to write using the array features of the 'numarray' packages of NumPy. The other program solves optimization problems with Differential Evolution (DE), and was a bit more difficult to write. I'll try to polish this DE code tomorrow. In any case, both codes are about as readable as the Fortran 90/95 versions I based them.

However, Python has both good and bad points. The implicit typing of the variables posed suprises, when an array was of integer type when I though it would be floating-point. The interactive style of programming was nice to have when debugging. I also liked the easy was to document modules and functions in Python.

There are still a lot of things to learn. I didn't much play with processing text files, for example. And programming graphical interfaces is way too vague for me at the moment. But perhaps I'll have the time to try these things as well.