new cmb code... (fwd)

[szajtan]

ok. ciakawa koncepcja ale nie do konca rozumiem jak to dziala.
masz obliczona i zparametryzowana (parametrami) w jakis sposob siatke modeli
(o roznych parametrach kosmologicznych).
tak ze na podstawie zadanych wartosci parametrow (nie kosmolog.) mozna
szybko odtworzyc widmo.
poszukiwane widmo w n wymiarowej przestrzeni parametrow jest otoczone przez
2^n modeli bazowych (o znanych zestawach wartosci parametrow
kosmologicznych, i tych z parametryzacji). Mozna tez obliczyc jak daleko w
przestrzeni parametrow kosmologicznych lezy poszukiwany model od modelow
budujacych wezly siatki (tylko jak z tego skorzystac). Wiemy rowniez o ile
roznia sie miedzy soba zestawy wartosci parametrow przypisanych do wezlow
sieci. Ale nie chwytam kewstii obliczenia poszukiwanych wartosci parametrow
na podstawie ktorych mozna odtworzyc widmo. Czyli innymi slowy jak to
interpolowac skoro wiemy tylko ze poszukiwany zestaw parametrow lezy gdzies
w oczku siatki parametrow.?


> I have the code - not publicly available yet, but soon
> it works as follows:
>
> 1. grid is produced and in nodes exact (well... cmbfast) cl spectrum is
> calculated
> 2. for every node some parametrisation of the spectrum is done (scaling,
> chebyshev coeficients etc.)
> 3. for given cosmo parms those parameters are interpolated and the
> spectrum is reconstructed.
>
> quite simple. but faster than other interpolation methods (for gridded
> models).
> that is the main idea...
>
> polarisation also possible, but not yet implemented.
> I can send you the code with the spectrum database if you wish.
>
> Bartosz Lew wrote:
>
> >ok, what's the catch ?
> >and where is the code ?
> >
> >and btw. how this works ?
> >
> >On Mon, 17 Feb 2003, Michal Frackowiak wrote:
> >
> >
> >
> >>well, not completed, not even documented, but if anyone of you would
> >>like to use it to reconstruct cmb spectrum REALLY FAST (within
> >>miliseconds) - please try.
> >>http://coyote.ncac.torun.pl/~michalf/cmbrecon/cmbrecon.php
> >>
> >>well - you can treat it as a preview of what is really about to happen.
> >>a fast cmb-spectrum reconstruction code. with quintessence (or whatever)
> >>included (soon also other scalar-tensor things). for generating grid:
> >>server-client architecture, quite smart reconstruction algorithms....
> >>in case of interest/doubts/will to use - feel free to contact me. code
> >>is quite complete and usable.
> >>
> >>michal frackowiak
> >>
> >>
> >>
> >>
> >>
> >
> >
> >
> >------------------------------------------------------------------------
> >
> >LISTNAME: shape-univ
> >HELP: send an email to sympa w astro.uni.torun.pl with "help"
> >WEB ARCHIVE: http://www.astro.uni.torun.pl/sympa/shape-univ/
> >UNSUBSCRIBE: email to sympa w astro.uni.torun.pl with "unsubscribe
shape-univ"
> >
> >
> >
> >
>
>
>
>


----------------------------------------------------------------------------
----


LISTNAME: shape-univ
HELP: send an email to sympa w astro.uni.torun.pl with "help"
WEB ARCHIVE: http://www.astro.uni.torun.pl/sympa/shape-univ/
UNSUBSCRIBE: email to sympa w astro.uni.torun.pl with "unsubscribe shape-univ"




--------------r-e-k-l-a-m-a-----------------


Tanie bilety lotnicze!
http://samoloty.onet.pl