diff --git a/joss-paper/paper.md b/joss-paper/paper.md index b938b80..9bcaaeb 100644 --- a/joss-paper/paper.md +++ b/joss-paper/paper.md @@ -13,16 +13,16 @@ authors: - name: Jonathan Pober equal-contrib: true # (This is how you can denote equal contributions between multiple authors) affiliation: 2 - orcid: 0000-0000-0000-0000 + orcid: 0000-0002-3492-0433 - name: Matthew Kolopanis affiliation: 3 orcid: 0000-0002-2950-2974 affiliations: - name: Scuola Normale Superiore, Italy index: 1 - - name: Brown University, USA + - name: Department of Physics, Brown University, Providence, RI, USA index: 2 - - name: Arizona State University, USA + - name: School of Earth and Space Exploration, Arizona State University, Tempe, AZ, USA index: 3 date: 18 January 2024 bibliography: paper.bib @@ -36,16 +36,14 @@ low-frequency radio experiments, including the MWA [@mwa], LOFAR [@lofar], HERA and the SKA [@Pritchard2015]. 21cmSense is a Python package that provides a modular framework for calculating the sensitivity of these experiments, in order to enhance the process of their design. -This paper presents version 2 of 21cmSense, which has been re-written from the ground up +This paper presents version v2.0.0 of 21cmSense, which has been re-written from the ground up to be more modular and extensible, and to provide a more user-friendly interface -- as well as converting the well-used legacy package, presented in [@Pober2014] from Python 2 to 3. -21cmSense computes noise estimates under the framework of *map-making*, in which the -many baselines of an interferometer are binned into a UV grid before a Fourier Transform -over the frequency axis is performed. This is a common approach in the field, although -other approaches exist, such the delay-spectrum method [@Parsons2012]. -The full sensitivity calculation in the map-making approach is rather involved and -computationally expensive in its most general form [@fhd], however 21cmSense uses a few +21cmSense can compute sensitivity estimates for both map-making [@fhd] and +delay-spectrum [@Parsons2012] approaches to power-spectrum estimation. +The full sensitivity calculation is rather involved and +computationally expensive in its most general form, however 21cmSense uses a few key assumptions to accelerate the calculation: 1. The UV grid is chosen to have cells that are comparable to the instrument's beam size. @@ -56,11 +54,6 @@ key assumptions to accelerate the calculation: 2. We do not consider flagging of visibilities due to RFI and other systematics, which can complicate the propagation of uncertainties. -Beyond these assumptions, there is also the current limitation that 21cmSense computes -the sensitivity under the map-making framework. Nevertheless, the modularity included -in this new version provides a path forward to include delay-spectrum calculations in -the future. - Some of the key new features introduced in this version of 21cmSense include: 1. Simplified, modular library API: the calculation has been split into modules that can @@ -76,7 +69,7 @@ Some of the key new features introduced in this version of 21cmSense include: 5. Generalization of the sensitivity calculation. The `Sensitivity` class is an abstract class from which the sensitivity of differing summary statistics can be defined. Currently, its only implementation is the `PowerSpectrum` class, which computes the - classic sensitivity of the (map-making style) power spectrum. However, the framework + classic sensitivity of the power spectrum. However, the framework can be extended to other summaries, for example wavelets [@Trott2016a]. 6. Improved speed: the new version of 21cmSense is significantly faster than the legacy version, due to a number of vectorization improvements in the code.