If you think this could be interesting to others you can reply or cc
your opinion on shape-univ.
Daję odpowiedź chyba niedługi...
Czesc Boud.
I have a new idea.
I've just compiled the healpix soft and generated a map from
previously calculated power spectrum from new cmbfast package.
eg. see
http://adjani.astro.uni.torun.pl/~blew/praca/topology/moj-l04-08-anim.gif
or if it's too slow moj-l04-08-anim2.gif :))
I was just thinking about that tast of significance of the signal
you've found and what you said that this is a lot of work doesn's seem
to be right to me now. well imagine I provide you a new simulated
map. in fits format for standard LCDM WMAP model in WMAP resolution
and pixelisation scheme. you put this through you pileline for dodec just
to see
what comes out. the rms signal or peak-to-peak dispersion would define
the valid detection threshold for the real map. isn't that right ?
it doesn't matter that the simulation is performed for simply
connected space. - even better. in fact it's similar as doing the same
thing but on the realmap but with nonphysical case of zero rotation.
I'll try to do it tommorow on my own with your package if I'll manage
to find out how it works and compile it correctly.
what do you think. of couser to treat this seriously I must pay more
attention to noise thing in each pixel, size of beam etc but as I
remember I read some good publication about this stuff recently so
perhaps this isn't big problem.
the bigger problem could possibly arise when I'd like to simulate the
map for some non-trivial topology to get some strong signal in a shape
of spikes in circles correlation function, like cornish et al. did.
If you think this could be interesting to others you can reply or cc
your opinion on shape-univ.
bartek
Boud Roukema