2D probe of multi-parameter functions: an MCMC alternative
Joel S. Jayson
arXiv:2404.08718v1 Announce Type: new
Abstract: We evaluate multi-parameter functions by fixing all parameter values, but two. The temperature power spectrum of the lensed cosmic microwave background (CMB) serves as an example in which the amplitude, $A_{s}$, nears linearity at small deviations, reducing power spectra computation to a 1D problem. We apply the analysis to assess the shift in the apparent value of ${H_0}$ as a result of cosmic infrared background (CIB), and Poisson point sources (PS). By iteratively cycling through parameters paired with $A_{s}$, and within a few hundred calls for spectra, we derive values in agreement with Planck. We adjudge that when neither variable is linear, thousands of calls are required, still competitive with the MCMC method.arXiv:2404.08718v1 Announce Type: new
Abstract: We evaluate multi-parameter functions by fixing all parameter values, but two. The temperature power spectrum of the lensed cosmic microwave background (CMB) serves as an example in which the amplitude, $A_{s}$, nears linearity at small deviations, reducing power spectra computation to a 1D problem. We apply the analysis to assess the shift in the apparent value of ${H_0}$ as a result of cosmic infrared background (CIB), and Poisson point sources (PS). By iteratively cycling through parameters paired with $A_{s}$, and within a few hundred calls for spectra, we derive values in agreement with Planck. We adjudge that when neither variable is linear, thousands of calls are required, still competitive with the MCMC method.

Comments are closed, but trackbacks and pingbacks are open.