carronj
commented
1 year ago All this makes perfect sense to me.
Le 15 mars 2023 à 09:20, mreineck @.***> a écrit :
I compared the ALM2MAP_DERIV SHT mode to the spin-SHTs, and as far as I can tell, it is identical to a spin-1 SHT where the curl component is zero, with one difference: the input a_lm are multiplied with an additional factor sqrt(l*(l+1)).
Does that agree with your observations, @carronj https://github.com/carronj ?
If that is indeed the case (and if the transform works in an analogous fashion for higher spins), introducing grad-only alm2map and adjoint_alm2map should not be a big deal. In this mode I'd like to omit the sqrt(l*(l+1)) factor however,and the user should apply it if needed.
— Reply to this email directly, view it on GitHub https://github.com/mreineck/ducc/issues/11, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACW2IOLJVX5XY52V22BPAJTW4F3UXANCNFSM6AAAAAAV3OVQ5A. You are receiving this because you were mentioned.
mreineck
I compared the
ALM2MAP_DERIVSHT mode to the spin-SHTs, and as far as I can tell, it is identical to a spin-1 SHT where the curl component is zero, with one difference: the inputa_lmare multiplied with an additional factorsqrt(l*(l+1)).Does that agree with your observations, @carronj ?
If that is indeed the case (and if the transform works in an analogous fashion for higher spins), introducing grad-only
alm2mapandadjoint_alm2mapshould not be a big deal. In this mode I'd like to omit thesqrt(l*(l+1))factor however,and the user should apply it if needed.