[GenABEL-dev] faster polygenic

[Yurii Aulchenko]

Hi Gulya,

As I mentioned, I am also very bothered by the fact that 'polygenic'
works so slowly. I cc to Xia who mentioned he had some idea on how to
make it working faster -- Xia, would be great if you could contribute
to the discussion! I also cc to GA devel list, so people know what is
going on.

Here I will describe how polygenic works currently (very far from optimal!)

'polygenic'

https://r-forge.r-project.org/scm/viewvc.php/pkg/GenABEL/R/polygenic.R?view=markup&revision=623&root=genabel

is numerically maximizing likelihood function computed by procedure 'polylik'

https://r-forge.r-project.org/scm/viewvc.php/pkg/GenABEL/R/polylik.R?view=markup&revision=545&root=genabel

For maximization, standard R procedures 'nlm' or 'optim' can be used
-- see what happens in 'polygenic' code between lines 246-300.

Any ideas on how we can speed it up are very welcome!

Yurii