Applications

X-PSI has been applied in the following studies. In addition to giving an idea of the scientific applications, these may also be useful for performance benchmarking and planning of computational resource consumption.

If you have used X-PSI for a project and would like to link it here, please contact the X-PSI team and/or submit a pull-request on GitHub.

Early X-PSI development

X-PSI was initiated by Thomas Riley as part of his PhD work at the University of Amsterdam. Chapter 3 of his PhD thesis (Riley 2019) provides an extended technical description of v0.1 of X-PSI and contains results of some parameter recovery tests using synthetic data.

NICER papers

X-PSI has been used in several Pulse Profile Modeling analysis papers by the NICER team. These are typically published with a Zenodo repository containing data, model files, X-PSI scripts, samples and post-processing notebooks to enable full reproduction and independent analysis.

Salmi et al. 2022 (ApJ, 941, 150) used v0.7.10 of X-PSI to model NICER observations of the rotation-powered millisecond X-ray pulsar PSR J0740+6620 using NICER background estimates. See also the associated Zenodo repository.

Riley et al. 2021 (ApJL, 918, L27) used v0.7.6 of X-PSI to model NICER observations of the rotation-powered millisecond X-ray pulsar PSR J0740+6620. See also the associated Zenodo repository.

Bogdanov et al. 2021 (ApJL, 914, L15) provides additional details of the models used for NICER Pulse Profile Modeling, reports ray-tracing cross-tests for X-PSI and other codes for very compact stars where multiple imaging is important, and reports some parameter recovery simulations for synthetic data.

Bogdanov et al. 2019 (ApJL, 887, L26) reports the results of ray-tracing cross-tests for several codes in use in the NICER team including X-PSI.

Riley et al. 2019 (ApJL, 887, L21) used v0.1 of X-PSI to model NICER observations of the rotation-powered millisecond X-ray pulsar PSR J0030+0451. See also the associated Zenodo repository.

Below we give details of some settings for this paper. For later papers some changes were made - see papers for details.

Resource consumption: The calculations reported were performed on the Dutch national supercomputer Cartesius, mostly on the Broadwell nodes (i.e., using CPUs only, usually Broadwell architecture). A number of models were applied, with increasing complexity. Typically, ~1000+ cores were used for posterior sampling, when the parallelisation is weighted by time consumed. In total, the sampling problems in R19 required ~450,000 core hours.

Likelihood function settings: On compute nodes, the likelihood function evaluation times ranged from ~1 to ~3 seconds (single-threaded), depending on the model and point in parameter space.[1] The parallelisation was purely via MPI, with all but one process receiving likelihood function evaluation requests.

Resolution settings: Number of surface cells/elements per hot (closed) region:[2] 24x24; Number of phases (linearly spaced) = 100; Number of energies (logarithmically spaced) = 175; Number of rays (per parallel; linear in cosine of ray angle alpha):[3] 200

Interpolation settings: Steffen splines (GSL) were used everywhere apart from for the atmosphere, and for the atmosphere cubic polynomial interpolants were used in four dimensions.

Compilation: Intel compiler collection,[4] with the CORE-AVX2 instruction set, for X-PSI and dependencies (apart from numpy, which was centrally installed).

Footnotes

[1]	Variation between models on the same processor results from a combination of model complexity and resolution settings. Variation as a function of parameters also occurs due to issues like mesh construction on the surface subject to the resolution settings.

[2]

This is the approximate number of cells that have centres lying within a hot region. For models where a hot region has two temperature components these cells are split between the subregions according to area only such that the surface density of cells is commensurate—at least for subregions that are commensurate in shape. The fast precomputation mode was deactivated but if it were activated the distribution of cells would further be weighted by the approximate total number of counts generated by each subregion.

[3]	The rays were integrated for each likelihood function call instead of loading lookup tables from disk, or using an analytic treatment such as a high-order expansion (see, e.g., rayXpanda). Only primary images were included.

[4]	On Cartesius (and Lisa), one can simply execute module load intel/2017b to access these compilers from the toolchain.