Healpix2.0 map2alm

Use of Healpix, camb, CLASS, cosmomc, compilers, etc.
Post Reply
Rockhee Sung
Posts: 5
Joined: November 13 2007
Affiliation: University of Cape Town

Healpix2.0 map2alm

Post by Rockhee Sung » November 13 2007

I'm trying to have alms(T,E,B) from map with map2alm,Healpix subroutine.

For simple testing, I put alm=0 except a2,1=1 for temperature and got the proper map but reproduced alms from map2alm,healpix subroutine, haven't got same values like input ones. They have even imaginary part!

Simply speaking,

Input alm =0 or 1
(alm2map) -> Output map ->(map2alm) ->
Output alm = a+bi (nonzero b)

Is there anyone who has a idea?
It seems that imaginary part from ring_analysis for m are nonzero component but still don't know what's wrong with them..
For detail, I used w8ring_TQU=1.d0
zbounds=/-1.d0,1.d0/, Nside=8, nlmax=16

Martin Reinecke
Posts: 13
Joined: July 26 2006
Affiliation: Max-Planck-Institute for Astrophysics

Healpix2.0 map2alm

Post by Martin Reinecke » November 13 2007

The map -> a_lm step is not exact for the Healpix pixelisation; this is the case for many sperical grids in practical use.

For your setup I would expect that you should obtain a_lm that differ from the original input by up to 0.001. This can be improved by using anafast's iterative analysis mode.

Post Reply