Optimized Large-Scale CMB Likelihood And Quadratic Maximum Likelihood Power Spectrum Estimation
arXiv:1506.04273 · doi:10.1088/0067-0049/221/1/5
Abstract
We revisit the problem of exact CMB likelihood and power spectrum estimation with the goal of minimizing computational cost through linear compression. This idea was originally proposed for CMB purposes by Tegmark et al.\ (1997), and here we develop it into a fully working computational framework for large-scale polarization analysis, adopting \WMAP\ as a worked example. We compare five different linear bases (pixel space, harmonic space, noise covariance eigenvectors, signal-to-noise covariance eigenvectors and signal-plus-noise covariance eigenvectors) in terms of compression efficiency, and find that the computationally most efficient basis is the signal-to-noise eigenvector basis, which is closely related to the Karhunen-Loeve and Principal Component transforms, in agreement with previous suggestions. For this basis, the information in 6836 unmasked \WMAP\ sky map pixels can be compressed into a smaller set of 3102 modes, with a maximum error increase of any single multipole of 3.8\% at , and a maximum shift in the mean values of a joint distribution of an amplitude--tilt model of 0.006. This compression reduces the computational cost of a single likelihood evaluation by a factor of 5, from 38 to 7.5 CPU seconds, and it also results in a more robust likelihood by implicitly regularizing nearly degenerate modes. Finally, we use the same compression framework to formulate a numerically stable and computationally efficient variation of the Quadratic Maximum Likelihood implementation that requires less than 3 GB of memory and 2 CPU minutes per iteration for , rendering low- QML CMB power spectrum analysis fully tractable on a standard laptop.
arXiv comment13 pages, 13 figures, accepted by ApJS
References in corpus (27)
- Planck 2013 results. XVI. Cosmological parameters
- HEALPix -- a Framework for High Resolution Discretization, and Fast Analysis of Data Distributed on the Sphere
- Nine-Year Wilkinson Microwave Anisotropy Probe (WMAP) Observations: Cosmological Parameter Results
- Cosmological parameters from CMB and other data: a Monte-Carlo approach
- Nine-Year Wilkinson Microwave Anisotropy Probe (WMAP) Observations: Final Maps and Results
- Planck 2013 results. I. Overview of products and scientific results
- Three Year Wilkinson Microwave Anisotropy Probe (WMAP) Observations: Polarization Analysis
- An All-Sky Analysis of Polarization in the Microwave Background
- Karhunen-Loeve eigenvalue problems in cosmology: how should we tackle large data sets?
- MASTER of the CMB Anisotropy Power Spectrum: A Fast Method for Statistical Analysis of Large and Complex CMB Data Sets
- Estimating the Power Spectrum of the Cosmic Microwave Background
- How to measure CMB power spectra without losing information
- Planck 2013 results. XII. Component separation
- Component separation methods for the Planck mission
- Global, Exact Cosmic Microwave Background Data Analysis Using Gibbs Sampling
- Power spectrum estimation from high-resolution maps by Gibbs sampling
- Application of Monte Carlo Algorithms to the Bayesian Analysis of the Cosmic Microwave Background
- New constraints on Parity Symmetry from a re-analysis of the WMAP-7 low resolution power spectra
- Cosmological Parameter Constraints as Derived from the Wilkinson Microwave Anisotropy Probe Data via Gibbs Sampling and the Blackwell-Rao Estimator
- Low Variance at large scales of WMAP 9 year data
- New estimates of the CMB angular power spectra from the WMAP 5 yrs low resolution data
- The Challenge Of Data Analysis For Future CMB Observations
- Bayesian joint estimation of non-Gaussianity and the power spectrum
- A multi-level solver for Gaussian constrained CMB realizations
- Cosmological Parameters from a re-analysis of the WMAP-7 low resolution maps
- Grid-based exploration of cosmological parameter space with Snake
- CMB likelihood approximation for banded probability distributions
Cited by in corpus (8)
- BeyondPlanck I. Global Bayesian analysis of the Planck Low Frequency Instrument data
- Extreme data compression for the CMB
- BeyondPlanck XI. Bayesian CMB analysis with sample-based end-to-end error propagation
- BeyondPlanck XII. Cosmological parameter constraints with end-to-end error propagation
- ECLIPSE: a fast Quadratic Maximum Likelihood estimator for CMB intensity and polarization power spectra
- The XFaster Power Spectrum and Likelihood Estimator for the Analysis of Cosmic Microwave Background Maps
- Improved Gibbs samplers for Cosmic Microwave Background power spectrum estimation
- Robust Likelihoods for Inflationary Gravitational Waves from Maps of Cosmic Microwave Background Polarization