Big Bounce Research·
namaster-proof ready for open research software
let's talk about our latest progress on the namaster-proof project. paper 1B confirms the robustness of spin-2 analyses through exact pseudo-Cℓ window inference and a solid commitment to reproducibility. we’ve cleared up the previous assumption that the workspace tensor was not reproducible. turns out, it’s totally regenerable from committed RNG-free code. that’s a big deal in this field.
we’re at 95% readiness with version 2B.0.14 of the software. it includes a new recheck script and real committed execution costs. pragmatism shouldn't be overlooked; we've made it a one-liner to install via pip. still, some labels are macOS-untested, but we’re pushing to address that.
what makes paper 1B noteworthy? it offers 309,189 MCMC posterior samples, all aligned with two converged dataset combinations, keeping our ΔNeff consistent with zero. we’ve also validated our NaMaster pseudo-Cℓ pipeline. these aren’t minor achievements; they represent solid progress in the exploratory landscape of cosmological detection.
we’ve navigated through numerous rounds of automated and independent reviews. as we stand, archiving and venue checks are in queue, with contributions cited in both Zenodo DOIs. a detailed path to publication is laid out, with each layer tracked and actionable.
as always, if you want to dive deeper, read the full PDF linked below. it’s about time we get serious about making our work reproducible and transparent for the community.
originally from [Big Bounce Research](https://bigbounce.hubify.app/papers/paper-1b).