[ODE] Remote forces using ODE optimization

Sergey Kurdakov k15c20 at zelmail.ru
Fri Jun 25 02:53:28 MST 2004


Hello Kelvin,

as I have my ideas ;) second message from me  on the subject


now I would describe somewhat more easy approach to implement things which could help you .

so - could we think on that your suggestion

>For example is it possible to set a grid with a given size and ODE will
>give me the pair of proteins as before to allow me to optimize my remote
>forces calcul.

as a way to a solution.

 let us say - initially you consider  your world as some cube ( maybe of very big sizes...  hope you will have enough memory)
as  just a 3 dimentional array,  each element of this array  - array[i][j][k] will be  a list of proteins in the given 3D cell.

now I assume you could request pos of each protein and thus (esp if the grid is regular ( is splitted on equal intervals) you could
easily compute i j k where to place ( in a list of a respective cell) given protein

now - instead of the computing forces from all the proteins

 for any given protein you would be able to request all neighbouring
cells ( +- to i j and k I think the math to select near cells is streigforward) and then traverse their protein lists ( if they are
empty Ok no forces) if there are proteins you d compute forces...)

so this is a basic appproch I could think of. Maybe this will reduce your computational time.

Regards
Sergey



More information about the ODE mailing list