September 23, 2012

In ages...

Dear Readers,

I haven't written in ages... And yet so much has happened in my life! For starters, I've gotten married :-) So now it's two of us that make one.

Another thing I haven't done in ages: write a program, think of algorithms, use Scilab. I did all these three in the past week, after playing with this new popular toy called NeoCube / Buckyballs / Nanodots: they're really hard to put down and indeed they raise interesting questions of stable conformations in space.

In which directions are the poles of each ball aligned? That's what I want to know!

I looked up the magnetic field of a dipole, the potential energy of a dipole in a magnetic field and thought of how to implement the computation of stable conformations.
My script found the orientation of the poles in this 2D configuration...

For a start, I installed Scilab, which I hadn't used in years, and despite annoying bugs and crashes (version 5.3.3 wouldn't run on my Mac, I'm running 5.4.0-beta3) here are three results I eventually obtained that gave me comfort: I'm not entirely rusty yet and can still program something simple.

The next stages would be the have proper algorithms to compute stable conformations in 3D and I know that that is a difficult problem (e.g. protein folding...)


... and in this one (which I hadn't guessed)
And here, I'm sorry you can barely see the plot but it was really the first quantitative result of the algorithm that I couldn't have predicted with simple logic: at which angle exactly are the poles at the two bounds of this chain?