Table 1
Estimation of Fourier-Bessel coefficients: comparison of the new method, the reverse formulation (Eqs. (19) and (20) using HEALPix discretisation) with the original, raw formulation (Eq. (7)).
N | nside | ϵr(fℓmn) < 0.3% | ϵr(Cℓn) < 0.3% | T | |
[0, 20]/[21, 50] | [0, 20]/[21, 50] | ||||
|
|||||
4.2e5 | 512 | 87%/42% | 99%/96% | 8 | |
1024 | 95%/65% | 99%/98% | 4 | ||
2048 | 99%/84% | 99%/99% | 2 | ||
3.1e6 | 512 | 92%/50% | 99%/95% | 10 | |
1024 | 98%/74% | 100%/100% | 5 | ||
2048 | 99%/90% | 100%/100% | 2 | ||
9.7e6 | 512 | 92%/50% | 100%/97% | 12 | |
1024 | 97%/74% | 100%/100% | 6 | ||
2048 | 99%/90% | 100%/100% | 3 | ||
|
|||||
Notes. The third (fourth) column gives the percentage f coefficients for which the relative accuracy ϵ(fℓmn) (ϵ(Cℓn)) is lower than 0.3% for given values of nside and N. T is the ratio of elapsed times of the two methods.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.