Why it exists
Astrophysical interpretations of the pulsar-timing-array background require more than a smooth power law. A finite cosmic population of supermassive black-hole binaries produces realization variance, individually bright systems, anisotropy, and potentially non-Gaussian statistics. Studying those effects requires generating many internally consistent populations rather than repeatedly assembling one-off simulation scripts.
fastropop provides a compact, JAX-first implementation of a semi-analytic population model, together with the transformations needed to turn sampled binaries into PTA-frequency spectra and sky maps. Vectorized sampling, batched calculations, just-in-time compilation, and accelerator-compatible operations are used so that repeated population realizations do not become a bottleneck themselves.
What it provides
- A semi-analytic model of the cosmic supermassive black-hole binary population.
- Characteristic-strain quantities and expected binary counts.
- Reproducible, batched JAX-backed population realizations.
- Efficient binning into PTA-style frequency spectra.
- HEALPix gravitational-wave skymaps through
jax-healpyorhealpy.
Why speed matters
A single synthetic population is rarely enough: realization variance, rare bright binaries, anisotropy, and parameter dependence only become visible across ensembles. fastropop is designed to replace slow Python-level loops and fragmented one-off pipelines with a high-throughput workflow, making those ensemble studies and interactive parameter scans practical. The precise gain depends on the model, hardware, batch size, and whether compilation is included.
Who it is for
The package is useful when population-to-population fluctuations are part of the scientific question: testing Gaussian-background assumptions, studying bright-source contributions, producing synthetic sky realizations, or comparing astrophysical populations with PTA observables. It is currently an early-stage research package, and its documented model assumptions and validation tests remain important when interpreting the speed and outputs.
Research context
fastropop was developed with Chiara Cecchini, Gabriele Franciolini, and Mauro Pieroni, and was used in the analysis Are PTA measurements sensitive to gravitational-wave non-Gaussianities?