Iterative map-making with two-level preconditioning for polarized cosmic microwave background data sets - A worked example for ground-based experiments

Free Access

Table 2.

Median, 1σ statistics of residual norms, ||r̂^(k) || and computational cost of PCG runs for different scanning strategies.

		Median	Percentile	kcpuh
SP	M_BD	3 × 10⁻⁵	0.0001	8.1^†
	M_2l, ϵ_Arn = 10⁻⁶	9 × 10⁻⁶	3 × 10⁻⁵	12.6^†
	M_2l simpl.	7 × 10⁻⁷	2 × 10⁻⁶	10^†
	M_2l, ϵ_Arn = 10⁻¹²	1 × 10⁻⁷	3 × 10⁻⁶	20.8^†
LP	M_BD	0.001	0.0004	10.8
LP	M_2l simpl.	3 × 10⁻⁷	1 × 10⁻⁷	8.5
VLP	M_BD	10⁻⁵	10⁻⁵	19.2
VLP	M_2l simpl.	10⁻⁷	10⁻⁷	15.9

Notes. We consider p_16th and p_84th, the 16th and 84th percentiles, respectively, as 1σ upper and lower bounds. In the fourth column, we quote (p_84th − p_16th)/2.

^(†)

Values rescaled from EDISON to CORI computational system to better compare performances.

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.