Skip to content

References and provenance

Where each idea or piece of code in virgil came from: a paper, a package or a person. This page is generated from references/provenance.toml, the single list that every docstring and docs page cites; BibTeX is in references/virgil.bib. The Contributions page tells the history of the project in prose.

"Ported" means code was adapted; "after" means the idea and its equations were reimplemented from the paper or the package's documentation, with no code copied. An entry is marked verified when its arXiv ID or DOI was checked against the arXiv abstract page, Crossref or NASA ADS. Entries marked "Not yet verified" have not had that check: treat their identifiers and attributions as unconfirmed until the mark is removed.

Origin of the package

  • Functions for the PDS 70 JWST/AMI analysis by Dori Blakely (Blakely et al. 2024), rewritten by Benjamin Pope and Louis Desdoigts as drpangloss. Louis Desdoigts wrote the JAX grid-search syntax and the first kernel-phase support; everything is built on his zodiax (the dot-path fitting interface). The name is from Voltaire's Candide, after CANDID (Gallenne et al. 2015).
  • Jonah Goldfine: the acronym, and the parametric scenes in virgil.scenes (from his frito).
  • Toon De Prins: ModulatedGaussianRim (an azimuthally modulated thin ring times a Gaussian envelope) and the elliptical (u, v) transform helpers, from his post-AGB disk work. The literature for such rim models is Lazareff et al. (2017) and Hillen et al. (2016).
  • Grid searches (virgil.grid_fit): Dori Blakely and Louis Desdoigts, in the drpangloss era. Only linear_flux_grid notes its equivalence to fouriever's lincmap, as a comparison.
  • Shashank Dholakia: the gravity-darkened star (equations from Espinosa Lara & Rieutord 2011; the implementation is Aaron Dotter's GDit, ported to JAX in his jax-interferometry), the spherical-harmonic star (harmonix), the Bessel functions (CEPHES, via harmonix, now jaxbessel) and spectrally dispersed data.
  • Max Charles: the image reconstruction in virgil.imaging is a port and extension of the visibility part of dorito, with ideas from Ian Czekala's MPoL.
  • Louis Desdoigts: virgil.amigo reads the mixed-DISCO format of AMIGO; virgil.coverage follows its latent visibility basis; the matrix Fourier transform was checked against dLux.

Sources by module

Each row is a place where code in virgil follows a paper or package. ported: code was adapted. after: the idea and equations were reimplemented from the paper or the package's documentation, with no code copied.

virgil._closure

Symbol Source Relation
ClosureNoise Kammerer et al. (2020), arXiv:2011.01209, ADS after

virgil._elr

Symbol Source Relation
(module) Espinosa Lara & Rieutord (2011), arXiv:1109.3038, ADS, doi:10.1051/0004-6361/201117252 after
(module) Dotter, GitHub ported
(module) Dholakia, GitHub ported

virgil._flat

Symbol Source Relation
(module) Hogg, Myers & Bovy (2010), arXiv:1008.4146, ADS after

virgil._gauss_newton

Symbol Source Relation
_solve_free Marquardt (1963), doi:10.1137/0111030 after
_solve_free Moré (1978), doi:10.1007/BFb0067700 after
rounding_dtol Higham (2002), doi:10.1137/1.9780898718027 after

virgil._geometry

Symbol Source Relation
grid_visibilities Desdoigts et al. (2024), arXiv:2406.08703, ADS, GitHub after
grid_visibilities Soummer et al. (2007), arXiv:0711.0368, ADS after

virgil._linear

Symbol Source Relation
(module) Luger, Foreman-Mackey & Hogg (2017), arXiv:1710.11136, ADS, doi:10.3847/2515-5172/aa96b5 after

virgil._nufft_spread

Symbol Source Relation
evaluate Barnett et al. (2019), arXiv:1808.06736, ADS, doi:10.1137/18M120885X after
evaluate Lee & Greengard (2005), ADS, doi:10.1016/j.jcp.2004.12.004 after
spread Oudoumanessah & Iollo, GitHub after

virgil._orbit_coordinates

Symbol Source Relation
anchor_unit_orbit Broucke & Cefola (1972), ADS, doi:10.1007/BF01228432 after
sample_z Moran (1975), doi:10.1017/s0021900200047720 after
ti_to_z Goldstein, Poole & Safko (2002) after
z_log_prior Miles (1965), doi:10.1093/biomet/52.3-4.636 after
zeta_to_eccentricity Ford (2006), arXiv:astro-ph/0512634, ADS, doi:10.1086/500802 after

virgil.aliases

Symbol Source Relation
(module) Jeffreys (1946), ADS, doi:10.1098/rspa.1946.0056 after
(module) Lucy (2014), arXiv:1309.2868, ADS after
(module) MacKay (1992), doi:10.1162/neco.1992.4.3.415 after
(module) Price-Whelan et al. (2017), arXiv:1610.07602, ADS after
(module) VanderPlas (2017), arXiv:1703.09824, ADS after
fit_orbit_aliases virgil (own design) (2026) after

virgil.amigo

Symbol Source Relation
load_oi_data Desdoigts et al. (2025), arXiv:2510.09806, ADS after

virgil.angles

Symbol Source Relation
AngleVector Thompson et al. (2023), arXiv:2402.01971, ADS, doi:10.3847/1538-3881/acf5cc, GitHub after

virgil.arcs

Symbol Source Relation
(module) virgil (own design) (2026) after

virgil.coverage

Symbol Source Relation
ami_grid_record Desdoigts et al. (2025), arXiv:2510.09806, ADS after
mask_transfer Sivaramakrishnan et al. (2022), arXiv:2210.17434, ADS after

virgil.detection

Symbol Source Relation
(module) Gross & Vitells (2010), arXiv:1005.1891, ADS after
DetectionMC Fawcett (2006), ADS, doi:10.1016/j.patrec.2005.10.010 after
bootstrap_null Efron (1979), doi:10.1214/aos/1176344552 after
local_nsigma Chernoff (1954), ADS, doi:10.1214/aoms/1177728725 after
local_nsigma Wilks (1938), doi:10.1214/aoms/1177732360 after
rescale_errors Kammerer et al. (2020), arXiv:2011.01209, ADS after

virgil.ensemble

Symbol Source Relation
combine Drevon et al. (2026), arXiv:2609.15365, ADS after
ensemble Drevon et al. (2026), arXiv:2609.15365, ADS after

virgil.epochs

Symbol Source Relation
marginal_loglike MacKay (1992), doi:10.1162/neco.1992.4.3.415 after

virgil.fields

Symbol Source Relation
(module) Enßlin, Frommert & Kitaura (2008), arXiv:0806.3474, ADS after
GaussianField Arras et al. (2020), arXiv:2002.05218, ADS after
GaussianField Edenhofer et al. (2024), arXiv:2402.16683, ADS after
GaussianField Enßlin (2018), arXiv:1804.03350, ADS after
GaussianField Lindgren, Rue & Lindström (2011), doi:10.1111/j.1467-9868.2011.00777.x after
GaussianField Enßlin group (Arras, Edenhofer et al.), link after
GaussianField Strang (1999), ADS, doi:10.1137/S0036144598336745 after
MultipoleField virgil (own design) (2026) after
field_spectrum Lindgren, Rue & Lindström (2011), doi:10.1111/j.1467-9868.2011.00777.x after
field_spectrum Strang (1999), ADS, doi:10.1137/S0036144598336745 after

virgil.fitting

Symbol Source Relation
(module) Kingma & Ba (2014), arXiv:1412.6980, ADS after
(module) Liu & Nocedal (1989), ADS, doi:10.1007/BF01589116 after
fit Byrd, Lu, Nocedal & Zhu (1995), ADS, doi:10.1137/0916069 after
fit Charles et al. (2026), arXiv:2510.10924, ADS ported
fit Hutchinson (1989), ADS, doi:10.1080/03610918908812806 after
fit Skilling & Bryan (1984), ADS after
fit Zhu, Byrd, Lu & Nocedal (1997), doi:10.1145/279232.279236 after
gauss_newton_diag Bekas, Kokiopoulou & Saad (2007), ADS, doi:10.1016/j.apnum.2007.01.003 after
gauss_newton_diag Cornwell & Evans (1985), ADS after
gauss_newton_diag Greer, GitHub after
gauss_newton_diag Hutchinson (1989), ADS, doi:10.1080/03610918908812806 after
gauss_newton_diag Martens (2010), link after
gauss_newton_diag Schraudolph (2002), doi:10.1162/08997660260028683 after
gauss_newton_diag Skilling & Bryan (1984), ADS after

virgil.gains

Symbol Source Relation
ClosureOffsets Lachaume (2021), arXiv:2104.07082, ADS, doi:10.1017/pasa.2021.20 after
ClosureOffsets Meimon, Mugnier & Le Besnerais (2005), ADS, doi:10.1364/JOSAA.22.002348 after
GainModes Lachaume (2021), arXiv:2104.07082, ADS, doi:10.1017/pasa.2021.20 after

virgil.grid_fit

Symbol Source Relation
(module) Jeffreys (1946), ADS, doi:10.1098/rspa.1946.0056 after
_flux_converged Higham (2002), doi:10.1137/1.9780898718027 after
linear_flux_grid Gallenne et al. (2015), arXiv:1505.02715, ADS after
linear_flux_grid Luger, Foreman-Mackey & Hogg (2017), arXiv:1710.11136, ADS, doi:10.3847/2515-5172/aa96b5 after

virgil.harmonics

Symbol Source Relation
from_image_harmonics Dumitrescu (2007), doi:10.1007/978-1-4020-5125-8 after
reflection_coefficients Barndorff-Nielsen & Schou (1973), doi:10.1016/0047-259X(73)90030-4 after

virgil.imaging

Symbol Source Relation
(module) Tiede et al. (2025), arXiv:2511.17706, ADS after
Image Charles et al. (2026), arXiv:2510.10924, ADS ported
LCurve.classic_maxent Gull (1989), doi:10.1007/978-94-015-7860-8_4 after
LCurve.classic_maxent Skilling (1989), doi:10.1007/978-94-015-7860-8_3 after
LCurve.dof_discrepancy Craven & Wahba (1979), doi:10.1007/BF01404567 after
LCurve.dof_discrepancy Hutchinson (1989), ADS, doi:10.1080/03610918908812806 after
LCurve.dof_discrepancy Morozov (1966) after
LCurve.dof_discrepancy Wahba (1983), doi:10.1111/j.2517-6161.1983.tb01239.x after
Laplacian Thiébaut & Young (2017), arXiv:1708.08390, ADS after
LogSum Baron, Monnier & Kloppenburg (2010), ADS, doi:10.1117/12.857364 after
LogSum Candès, Wakin & Boyd (2007), arXiv:0711.1612, ADS after
MaxEntropy Baron & Young (2008), ADS, doi:10.1117/12.789115 after
MaxEntropy Skilling (1989), doi:10.1007/978-94-015-7860-8_3 after
MaxEntropy Skilling & Bryan (1984), ADS after
StarletL1 Akiyama et al. (2017), arXiv:1702.07361, ADS after
StarletL1 Honma, Akiyama, Uemura & Ikeda (2014), arXiv:1407.2422, ADS, doi:10.1093/pasj/psu070 after
TSV Akiyama et al. (2017), arXiv:1702.07361, ADS after
TSV Chael et al. (2018), arXiv:1803.07088, ADS, doi:10.3847/1538-4357/aab6a8, GitHub after
TSV Kuramochi et al. (2018), arXiv:1802.05783, ADS, doi:10.3847/1538-4357/aab6b5 after
TSV Rudin, Osher & Fatemi (1992), ADS, doi:10.1016/0167-2789(92)90242-F after
TV Chael et al. (2018), arXiv:1803.07088, ADS, doi:10.3847/1538-4357/aab6a8, GitHub after
TV Rudin, Osher & Fatemi (1992), ADS, doi:10.1016/0167-2789(92)90242-F after
clean Cornwell (2008), arXiv:0806.2228, ADS, doi:10.1109/jstsp.2008.2006388 after
clean Högbom (1974), ADS after
clean Rau & Cornwell (2011), arXiv:1106.2745, ADS, doi:10.1051/0004-6361/201117104 after
diagnose Thiébaut & Young (2017), arXiv:1708.08390, ADS after
error_scale MacKay (1992), doi:10.1162/neco.1992.4.3.415 after
l_curve Hansen & O'Leary (1993), ADS, doi:10.1137/0914086 after
laplace_samples MacKay (1992), doi:10.1162/neco.1992.4.3.415 after
log_evidence MacKay (1992), doi:10.1162/neco.1992.4.3.415 after
short_baseline_residual Ireland, Monnier & Thureau (2006), arXiv:2007.00716, ADS, doi:10.1117/12.670940 after
starlet Starck, Murtagh & Fadili (2010) after
starting_image Charles et al. (2026), arXiv:2510.10924, ADS ported

virgil.inference

Symbol Source Relation
(module) Betancourt (2017), arXiv:1701.02434, ADS after
(module) Hoffman & Gelman (2011), arXiv:1111.4246, ADS after

virgil.legacy.oifits_implaneia

Symbol Source Relation
(module) Greenbaum et al. (2014), arXiv:1411.3446, ADS after
(module) Greenbaum, Sivaramakrishnan et al., GitHub ported

virgil.likelihood

Symbol Source Relation
group_outlier_mixture Jeffreys (1946), ADS, doi:10.1098/rspa.1946.0056 after
inflated_errors Hogg, Bovy & Lang (2010), arXiv:1008.4686, ADS after

virgil.limits

Symbol Source Relation
absil_limits Absil et al. (2011), arXiv:1110.1178, ADS after
injection_limits Gallenne et al. (2015), arXiv:1505.02715, ADS after
nsigma Absil et al. (2011), arXiv:1110.1178, ADS after
ruffio_upperlimit Ruffio et al. (2018), arXiv:1809.08261, ADS after

virgil.metrics

Symbol Source Relation
(module) Baron et al. (2012), arXiv:1207.7141, ADS after
(module) Malbet et al. (2010), arXiv:1007.4473, ADS after
lawson_sigma_over_peak Lawson, Cotton & Hummel (2004), ADS, doi:10.1117/12.550710 after
rms_convolved Cotton, Monnier & Baron (2008), ADS, doi:10.1117/12.788903 after
Symbol Source Relation
DefensiveMixture Hesterberg (1995), doi:10.1080/00401706.1995.10484303 after
importance_evidence Hesterberg (1995), doi:10.1080/00401706.1995.10484303 after
resample_move Cabezas et al. (2024), arXiv:2402.10797, GitHub after
resample_move Gilks & Berzuini (2001), doi:10.1111/1467-9868.00280 after
resample_move Hoffman & Gelman (2011), arXiv:1111.4246, ADS after
resample_move Neal (2011), arXiv:1206.1901 after

virgil.models

Symbol Source Relation
(module) Pope et al., GitHub ported
(module) Desdoigts, GitHub after
ArchimedeanSpiral Han et al. (2020), arXiv:2008.05834, doi:10.1093/mnras/staa2349 after
ArchimedeanSpiral Tuthill, Monnier & Danchi (1999), arXiv:astro-ph/9904092, doi:10.1038/19033 after
ArchimedeanSpiral Tuthill et al. (2008), arXiv:0712.2111, doi:10.1086/527286 after
Crescent Benkevitch et al. (2016), arXiv:1609.00055 after
Crescent Kamruddin & Dexter (2013), arXiv:1306.3226, doi:10.1093/mnras/stt1068 after
FlaredDisk Blakely et al. (2024), arXiv:2404.13032, ADS after
GravityDarkenedStar Espinosa Lara & Rieutord (2011), arXiv:1109.3038, ADS, doi:10.1051/0004-6361/201117252 after
Harmonics Barndorff-Nielsen & Schou (1973), doi:10.1016/0047-259X(73)90030-4 after
Harmonics Dumitrescu (2007), doi:10.1007/978-1-4020-5125-8 after
HarmonixModel Dholakia & Pope (2025), arXiv:2509.25433, ADS, GitHub after
HarmonixModel Luger et al. (2018), arXiv:1810.06559, ADS after
Image Czekala et al. (2025), arXiv:2502.00100, ADS, GitHub, link after
LimbDarkenedDisk Dholakia & Pope (2025), arXiv:2509.25433, ADS, GitHub after
LinearImage Akiyama et al. (2017), arXiv:1702.07361, ADS after
LinearImage Byrd, Lu, Nocedal & Zhu (1995), ADS, doi:10.1137/0916069 after
ModulatedGaussianRim Lazareff et al. (2016), arXiv:1611.08428, ADS after
MultipoleDisk Jennings et al. (2020), arXiv:2005.07709, ADS, GitHub after
MultipoleDisk Kennedy (2025), arXiv:2502.17619, ADS, GitHub after
MultipoleKernel Yoshida et al. (2026), arXiv:2607.12618, ADS, doi:10.3847/1538-4357/ae89ad, GitHub after
QuadraticLimbDarkenedDisk Kipping (2013), arXiv:1308.0009, ADS, doi:10.1093/mnras/stt1435 after
RadialGP Yoshida et al. (2026), arXiv:2607.12618, ADS, doi:10.3847/1538-4357/ae89ad, GitHub after
RadialGPFunction Yoshida et al. (2026), arXiv:2607.12618, ADS, doi:10.3847/1538-4357/ae89ad, GitHub after
SquareRootLimbDarkenedDisk Díaz-Cordovés & Giménez (1992), ADS after
SquareRootLimbDarkenedDisk Kipping (2013), arXiv:1308.0009, ADS, doi:10.1093/mnras/stt1435 after
System Kluska et al. (2014), arXiv:1403.3343, ADS after
cvis_binary Ireland (2013), arXiv:1301.6205, ADS after
cvis_binary Martinache (2010), arXiv:1009.3933, ADS, GitHub ported
cvis_limb_darkened_disk Quirrenbach et al. (1996), ADS after
multipole_disk_cvis Jennings et al. (2020), arXiv:2005.07709, ADS, GitHub after
multipole_disk_cvis Kennedy (2025), arXiv:2502.17619, ADS, GitHub after

virgil.night_summaries

Symbol Source Relation
_eps_rule Golub & Welsch (1969), doi:10.1090/S0025-5718-69-99647-1 after

virgil.observables

Symbol Source Relation
DifferentialPhase Luger, Foreman-Mackey & Hogg (2017), arXiv:1710.11136, ADS, doi:10.3847/2515-5172/aa96b5 after
FluxSpectrum Luger, Foreman-Mackey & Hogg (2017), arXiv:1710.11136, ADS, doi:10.3847/2515-5172/aa96b5 after

virgil.oidata

Symbol Source Relation
OIData.with_error_floor Mérand (2022), arXiv:2207.11047, ADS, GitHub after
OIData.with_north_angle Thompson et al. (2023), arXiv:2402.01971, ADS, doi:10.3847/1538-3881/acf5cc, GitHub after

virgil.oifits

Symbol Source Relation
read_oifits Duvert, Young & Hummel (2015), arXiv:1510.04556, ADS, doi:10.1051/0004-6361/201526405 after
read_oifits Pauls et al. (2005), arXiv:astro-ph/0508185, ADS, doi:10.1086/444523 after
write_oifits Duvert, Young & Hummel (2015), arXiv:1510.04556, ADS, doi:10.1051/0004-6361/201526405 after

virgil.orbit_cells

Symbol Source Relation
_Space.states_at Broucke & Cefola (1972), ADS, doi:10.1007/BF01228432 after
position_leg Gilks & Berzuini (2001), doi:10.1111/1467-9868.00280 after
position_leg Hesterberg (1995), doi:10.1080/00401706.1995.10484303 after
position_leg Hoffman & Gelman (2011), arXiv:1111.4246, ADS after
search_cell Fischler & Bolles (1981), doi:10.1145/358669.358692 after
search_cell Golub & Pereyra (1973), doi:10.1137/0710036 after
search_cell Hadi & Simonoff (1993), ADS, doi:10.1080/01621459.1993.10476407 after
search_cell Rousseeuw (1984), doi:10.1080/01621459.1984.10477105 after
search_cell Rousseeuw & Van Driessen (1999), doi:10.1080/00401706.1999.10485670 after
search_plan Fischler & Bolles (1981), doi:10.1145/358669.358692 after

virgil.orbits

Symbol Source Relation
KeplerOrbit Hartkopf, McAlister & Franz (1989), ADS, doi:10.1086/115193 after
KeplerOrbit Wright & Howard (2009), arXiv:0904.3725, ADS after
KeplerOrbit.from_position_angle Thompson et al. (2023), arXiv:2402.01971, ADS, doi:10.3847/1538-3881/acf5cc, GitHub after
PositionData.term Thompson et al. (2023), arXiv:2402.01971, ADS, doi:10.3847/1538-3881/acf5cc, GitHub after
_thiele_innes_fit Thiele (1883), ADS after
period_grid Lucy (2014), arXiv:1309.2868, ADS after
starting_orbits Lucy (2014), arXiv:1309.2868, ADS after
starting_orbits Luger, Foreman-Mackey & Hogg (2017), arXiv:1710.11136, ADS, doi:10.3847/2515-5172/aa96b5 after
starting_orbits Price-Whelan et al. (2017), arXiv:1610.07602, ADS after

virgil.pipeline._checks

Symbol Source Relation
r_hat Vehtari et al. (2021), arXiv:1903.08008, ADS after

virgil.pipeline.star

Symbol Source Relation
_compare Chernoff (1954), ADS, doi:10.1214/aoms/1177728725 after
_compare Self & Liang (1987), ADS, doi:10.1080/01621459.1987.10478472 after

virgil.priors

Symbol Source Relation
(module) Hogg, Myers & Bovy (2010), arXiv:1008.4146, ADS after
(module) Jeffreys (1946), ADS, doi:10.1098/rspa.1946.0056 after
hierarchical_scales Betancourt & Girolami (2013), arXiv:1312.0906, ADS after

virgil.radio

Symbol Source Relation
CalibrationNuisance Arras et al. (2019), arXiv:1903.11169, ADS after
CalibrationNuisance Blackburn et al. (2020), arXiv:1910.02062, ADS after
CalibrationNuisance Broderick et al. (2020), ADS, doi:10.3847/1538-4357/ab91a4 after
CalibrationNuisance Cornwell & Wilkinson (1981), ADS, doi:10.1093/mnras/196.4.1067 after
CalibrationNuisance Czekala et al. (2025), arXiv:2502.00100, ADS, GitHub, link after
CalibrationNuisance Zawadzki et al. (2023), arXiv:2209.11813, ADS, doi:10.1088/1538-3873/acdf84 after
ComplexVisData Pawitan (2001), doi:10.1093/oso/9780198507659.001.0001 after
ComplexVisData Zawadzki et al. (2023), arXiv:2209.11813, ADS, doi:10.1088/1538-3873/acdf84 after
ComplexVisData._imaging_weights Briggs (1995), ADS after
fit_staged_nuisances Cornwell & Wilkinson (1981), ADS, doi:10.1093/mnras/196.4.1067 after

virgil.radio_clean

Symbol Source Relation
_minor_cycle Clark (1980), ADS after
_minor_cycle Cornwell (2008), arXiv:0806.2228, ADS, doi:10.1109/jstsp.2008.2006388 after
clean Clark (1980), ADS after
clean Högbom (1974), ADS after
clean Offringa et al. (2014), arXiv:1407.1943, ADS after
clean Schwab (1984), ADS after

virgil.radio_toeplitz

Symbol Source Relation
ToeplitzVisData Barnett et al. (2019), arXiv:1808.06736, ADS, doi:10.1137/18M120885X after
ToeplitzVisData Fessler et al. (2005), ADS, doi:10.1109/TSP.2005.853152 after
ToeplitzVisData Lee & Greengard (2005), ADS, doi:10.1016/j.jcp.2004.12.004 after
ToeplitzVisData Oudoumanessah & Iollo, GitHub after

virgil.scenes

Symbol Source Relation
(module) Goldfine, GitHub ported

virgil.smc

Symbol Source Relation
bridge_smc Cabezas et al. (2024), arXiv:2402.10797, GitHub after
bridge_smc Du & Guyader (2021), arXiv:1909.13602, doi:10.1214/20-AAP1611 after
bridge_smc Lee & Whiteley (2018), arXiv:1509.00394, doi:10.1093/biomet/asy028 after
tempered_smc Cabezas et al. (2024), arXiv:2402.10797, GitHub after
tempered_smc Du & Guyader (2021), arXiv:1909.13602, doi:10.1214/20-AAP1611 after
tempered_smc Lee & Whiteley (2018), arXiv:1509.00394, doi:10.1093/biomet/asy028 after

virgil.spectra

Symbol Source Relation
(module) Hillen et al. (2016), arXiv:1603.03023, ADS, doi:10.1051/0004-6361/201628125 after
Nodes Mérand (2022), arXiv:2207.11047, ADS, GitHub after
PowerLaw Kluska et al. (2014), arXiv:1403.3343, ADS after

virgil.svi

Symbol Source Relation
_loss_settled Kass & Raftery (1995), doi:10.1080/01621459.1995.10476572 after
variational De Cao, Titov & Aziz (2019), arXiv:1904.04676, ADS after
variational Kingma et al. (2016), arXiv:1606.04934, ADS after
variational Kucukelbir et al. (2016), arXiv:1603.00788, ADS after
variational Vehtari et al. (2015), arXiv:1507.02646, ADS after

virgil.tempering

Symbol Source Relation
(module) virgil (own design) (2026) after

Compared, not ported

Studied for comparison; nothing was taken from them.

Source Note
Arras et al. (2020), arXiv:2008.11435, ADS A&A 646, A84 (2021)
Blunt et al. (2017), arXiv:1703.10653, ADS OFTI
Blunt et al. (2019), arXiv:1910.01756, ADS, GitHub orbitize!; AJ 159, 89 (2020)
Blunt et al. (2024), arXiv:2409.11573, ADS, doi:10.21105/joss.06756 orbitize!; JOSS 9, 6756
Brandt et al. (2021), arXiv:2105.11671, ADS, GitHub orvara; AJ 162, 186
CASA Team (2022), link Iteration Control, global stopping criteria: divergence 1 is (PeakRes - PrevPeakRes) / PrevPeakRes > 3.0, divergence 2 is (PeakRes - MinPeakRes) / MinPeakRes > 3.0, both over the whole image ignoring the mask (stop codes 5 and 6); minor cycle: (currentPeak - prevPeak) / prevPeak > 0.1, stop code 4 saying 'from minimum'. The page is inconsistent on the minor-cycle reference; virgil uses the running minimum per scale (virgil's choice). Read 2026-10-24
Frank, Leike & Enßlin (2021), arXiv:2105.10470, ADS geoVI; not implemented, svi.variational is a different method
Kammerer, GitHub comparison only: compare linear_flux_grid with its lincmap (closure/kernel phases only); no paper found
Knollmüller & Enßlin (2019), arXiv:1901.11033, ADS MGVI; laplace_samples shares its covariance form but is a single draw at the MAP, with no iteration
Mawet et al. (2014), arXiv:1407.2247, ADS ApJ 792, 97; the small-sample false-alarm correction, not applied in injection_recovery
Mérand et al., GitHub algorithm paper: gallenne2015
Selig et al. (2013), arXiv:1301.4499, ADS

Background literature

Cited in the documentation as motivation or terminology; not a source of virgil's methods.

Source Note
Abramowitz & Stegun (1964), link 25.4.38: Gauss-Chebyshev quadrature (first kind), exact for polynomials of degree 2n-1.
Bishop (2006) Springer textbook, section 3.5 (evidence approximation for regularization weights)
Born & Wolf (1999), doi:10.1017/CBO9781139644181 section 9.2: Zernike radial polynomials, R_n^m(r) ~ r^|m|, the regularity of a smooth field at the origin
Currie et al. (2019), arXiv:1905.04322, ADS, doi:10.3847/2041-8213/ab1b42 ApJL 877, L3; motivates fitting a disk and a planet together
Elsworth et al. (2025), arXiv:2508.15734, ADS Google's measured 0.24 Wh per median Gemini Apps text prompt; a cross-check for the Gemini figure in the development-carbon estimate (docs/dev_carbon.md)
Guardiani et al. (2022), arXiv:2105.13483, ADS PLoS ONE 17, e0275011; the Matérn amplitude that NIFTy.re cites (Edenhofer et al. 2024); virgil's MultipoleField is its own construction
Junklewitz et al. (2016), arXiv:1311.5282, ADS A&A 586, A76; an IFT imaging algorithm for radio interferometry
Kraus & Ireland (2012), arXiv:1110.3808, ADS, doi:10.1088/0004-637X/745/1/5 ApJ 745, 5; motivates fitting a disk and a planet together
Lannelongue, Grealey & Inouye (2021), ADS, doi:10.1002/advs.202100707 Advanced Science 8, 2100707; the method behind the development-carbon estimate (docs/dev_carbon.md)
Monnier et al. (2007), arXiv:0706.0867, ADS, doi:10.1126/science.1143205 Science 317, 342; the Altair-like truth of notebooks/mwe/mwe_elr_chara.ipynb is inspired by it, not a reproduction
Papaspiliopoulos, Roberts & Sköld (2007), ADS, doi:10.1214/088342307000000014 Statistical Science 22, 59; centred and non-centred parameterizations
Rasmussen & Williams (2006), doi:10.7551/mitpress/3206.001.0001, link MIT Press (book); GP terminology, Matérn kernels, marginal likelihood; A.9 matrix inversion and determinant lemmas (arrow_at)
Roettenbacher et al. (2016), arXiv:1709.10107, ADS, doi:10.1038/nature17444 Nature 533, 217; CHARA/MIRC starspot imaging that motivates HarmonixModel
Rue & Held (2005), doi:10.1201/9780203492024 Chapman & Hall/CRC (book)
Sallum et al. (2015), arXiv:1511.07456, ADS, doi:10.1038/nature15761 Nature 527, 342; motivates fitting a disk and a planet together
van Hamme (1993), ADS, doi:10.1086/116788 AJ 106, 2096; the quadratic law the square-root law is compared with

Other background sources

Heritage and context without a code symbol of their own.

Source Note
Eyer & Bartholdi (1999), arXiv:astro-ph/9808176, doi:10.1051/aas:1999102 A&AS 135, 1; for irregular sampling the usable Nyquist frequency can far exceed that of the mean gap: considered, and not used, for the census's sampling test (scripts/orbit_population.py, nyquist_ok)
Martinache (2010), arXiv:1009.3933, ADS ApJ 724, 464 (2010)
Modrák, Moon, Kim, Bürkner, Huurre, Faltejsková, Gelman & Vehtari (2023), arXiv:2211.02383 Bayesian Analysis 2023 (arXiv 2022); data-dependent test quantities are valid in SBC; the sky-plane rank of the population test (scripts/orbit_population.py)
Talts, Betancourt, Simpson, Vehtari & Gelman (2018), arXiv:1804.06788 rank-uniformity check of a posterior against its own prior draws; the population test of the cell fit (scripts/orbit_population.py)

In the bibliography only

Cited by the paper, not a source of virgil's methods.

Source Note
Albert (2020), arXiv:2012.15286 jaxns, the alternative JAX nested sampler considered for candidate (d); not measured, BlackJAX is already a dependency
Ben-Israel (1966), doi:10.1016/0022-247X(66)90115-6 J. Math. Anal. Appl. 15, 243; Newton with the Moore-Penrose inverse for underdetermined systems; planned minimum-norm set step and C-steps of the orbit-cells search
Beskos, Pillai, Roberts, Sanz-Serna & Stuart (2013), doi:10.3150/12-bej414 Bernoulli 19, 1501; acceptance target 0.651
Boyd & Vandenberghe (2004), doi:10.1017/CBO9780511804441 Cambridge University Press; 9.5.1: the Newton decrement as a stopping rule; planned stop of the orbit-cells set step and C-steps
Buchholz, Chopin & Jacob (2021), arXiv:1808.07730, ADS, doi:10.1214/20-ba1222 Bayesian Anal. 16, 745; fixed-L HMC kernels tuned per stage
Celeux & Govaert (1992), doi:10.1016/0167-9473(92)90042-E Comput. Stat. Data Anal. 14, 315; classification EM: block ascent over labels and parameters; planned inlier-scale step of the orbit-cells search
Chavez et al. (2026), arXiv:2606.04094, ADS
Chen, Feroz & Hobson (2023), arXiv:1908.04655, doi:10.1214/22-BA1323 Bayesian Analysis 18, 695; nested sampling with the prior replaced by any reference and the likelihood by prior x likelihood / reference, same Z; how NS is seeded by the search
Del Moral, Doucet & Jasra (2006), doi:10.1111/j.1467-9868.2006.00553.x JRSSB 68, 411; SMC samplers and the normalizing-constant estimate, Z as the product of the incremental weights
Foreman-Mackey et al. (2021), arXiv:2105.01994, ADS, doi:10.21105/joss.03285 JOSS 6, 3285
Fowlie, Handley & Su (2021), arXiv:2010.13884, doi:10.1093/mnras/stab590 MNRAS 503, 1199; likelihood plateaus break nested sampling's compression; the all-outlier floor is one
Gronau, Sarafoglou, Matzke, Ly, Boehm, Marsman, Leslie, Forster, Wagenmakers & Steingroever (2017), arXiv:1703.05984, doi:10.1016/j.jmp.2017.09.005 J. Math. Psychol. 81, 80; bridge sampling from MCMC draws to a fitted proposal (compared for the orbit alias-cell evidence)
Hattori et al. (2024), ADS, doi:10.5281/zenodo.10736936, GitHub Zenodo; https://github.com/exoplanet-dev/jaxoplanet
Jasra, Stephens, Doucet & Tsagaris (2011), doi:10.1111/j.1467-9469.2010.00723.x Scand. J. Stat. 38, 1 (online December 2010); ESS-adaptive tempering schedules
Kroupa & Handley (2026), arXiv:2602.22307 nested sampling on an aliased (time-delay) likelihood needs many live points to avoid spurious modes; the caution behind running NS as a check, not the default
Kustaanheimo & Stiefel (1965), doi:10.1515/crll.1965.218.204 J. reine angew. Math. 218, 204; regularizing maps of Kepler motion
Levi-Civita (1920), doi:10.1007/BF02404416 Acta Math. 42, 99; the planar regularizing map
Mantel & Haenszel (1959), doi:10.1093/jnci/22.4.719 J. Natl. Cancer Inst. 22, 719; stratified test of a common rate across strata; planned orbit population-test assertion
McDougall, Pope & Davis (2026), arXiv:2505.09832, doi:10.1017/pasa.2026.10149 PASA 43; hessian_scan: a scan over an aliased variable with optimization and a Laplace evidence per slice, integrated across the grid; prior art for the across-N propagation
Meng & Wong (1996), link Statistica Sinica 6, 831; bridge sampling and its optimal iterative estimator (compared for the orbit alias-cell evidence). No DOI, arXiv ID or ADS record: title and authors read from the journal page
Mengersen & Tweedie (1996), doi:10.1214/aos/1033066201 Ann. Stat. 24, 101; independence Metropolis-Hastings is uniformly ergodic iff the weight is bounded
Nocedal & Wright (2006), doi:10.1007/978-0-387-40065-5 Springer, 2nd ed.; ch. 12 and 16.7: KKT and projected-gradient conditions (kkt_report), KKT sign condition of a bound (fit's bound_stall); ch. 3-4 and 10.3: safeguarded steps in a bracket, trust-region predicted reduction, Newton and Levenberg-Marquardt in curved valleys; planned set step of the orbit-cells search
Phipson & Smyth (2010), doi:10.2202/1544-6115.1585 Stat. Appl. Genet. Mol. Biol. 9, 39; (1 + #)/(1 + n) permutation p-value
Price-Whelan et al. (2020), arXiv:2002.00014, ADS ApJ 895, 2; The Joker at scale
Ritter & Tanner (1992), doi:10.1080/01621459.1992.10475289 JASA 87, 861; griddy Gibbs: draw a scalar conditional by tabulating it on a grid (a review toy for the orbit alias-cell search)
Rue, Martino & Chopin (2009), doi:10.1111/j.1467-9868.2008.00700.x J. R. Stat. Soc. B 71, 319; INLA: quadrature in a low-dimensional parameter with conditional Laplace; planned ridge proposal component in rho for orbit cells
Shoemake (1992), doi:10.1016/b978-0-08-050755-2.50036-1 Graphics Gems III, 124; unit quaternions and the Haar measure of SO(3)
Skilling (2006), doi:10.1214/06-BA127 Bayesian Analysis 1, 833; nested sampling, candidate (d) for the per-cell evidence; robust to first-order phase transitions where tempering fails
Soulain et al., GitHub separate package, bibliography only; paper reference to add (SPIE 2020)
Surjanovic, Biron-Lattes, Tiede et al. (2023), arXiv:2308.09769, ADS parallel tempering prior art next to Octofitter
Tierney & Kadane (1986), doi:10.1080/01621459.1986.10478240 JASA 81, 82; Laplace's method for marginal likelihoods (the baseline for the orbit alias-cell evidence)
Veach & Guibas (1995), doi:10.1145/218380.218498 SIGGRAPH '95, 419; the balance heuristic, the partition of unity that splits the bridge per mode
Yallup, Kroupa & Handley (2026), arXiv:2601.23252 the BlackJAX nested slice sampler (blackjax.nss, >= 1.7) measured as candidate (d) on the leg-1 toys
Zhou, Johansen & Aston (2016), doi:10.1080/10618600.2015.1060885 JCGS 25, 701; conditional ESS for the next temperature

All sources

Alphabetical, one anchor per bibkey. Docstrings link here as ../references.md#<bibkey>.

  • Abramowitz & Stegun (1964). Handbook of Mathematical Functions. link. 25.4.38: Gauss-Chebyshev quadrature (first kind), exact for polynomials of degree 2n-1. Not yet verified against arXiv, Crossref or ADS.
  • Absil et al. (2011). Searching for faint companions with VLTI/PIONIER. I. Method and first results. arXiv:1110.1178, ADS.
  • Akiyama et al. (2017). Imaging the Schwarzschild-radius-scale Structure of M87 with the Event Horizon Telescope using Sparse Modeling. arXiv:1702.07361, ADS. ApJ 838, 1; SMILI: bounded linear (non-negative) pixels.
  • Albert (2020). JAXNS: a high-performance nested sampling package based on JAX. arXiv:2012.15286. jaxns, the alternative JAX nested sampler considered for candidate (d); not measured, BlackJAX is already a dependency.
  • Arras et al. (2019). Unified Radio Interferometric Calibration and Imaging with Joint Uncertainty Quantification. arXiv:1903.11169, ADS. A&A 627, A134 (2019); joint calibration and imaging (resolve), gain priors.
  • Arras et al. (2020). Comparison of classical and Bayesian imaging in radio interferometry: Cygnus A with CLEAN and resolve. arXiv:2008.11435, ADS. A&A 646, A84 (2021).
  • Arras et al. (2020). Variable structures in M87* from space, time and frequency resolved interferometry. arXiv:2002.05218, ADS. Nature Astron. 6, 259 (2022).
  • Barndorff-Nielsen & Schou (1973). On the parametrization of autoregressive models by partial autocorrelations. doi:10.1016/0047-259X(73)90030-4. J. Multivariate Anal. 3, 408-419; the unconstrained reflection-coefficient parametrization of a minimum-phase polynomial. Standard mathematics, credited here and in docstrings, not on the contributors page.
  • Barnett et al. (2019). A parallel nonuniform fast Fourier transform library based on an "exponential of semicircle" kernel. arXiv:1808.06736, ADS, doi:10.1137/18M120885X. SIAM J. Sci. Comput. 41, C479 (FINUFFT); the ES kernel, and ψ̂ at non-integer frequencies by quadrature (Sect. 3.3).
  • Baron & Young (2008). Image reconstruction at Cambridge University. ADS, doi:10.1117/12.789115. BSMEM; SPIE 7013.
  • Baron et al. (2012). The 2012 Interferometric Imaging Beauty Contest. arXiv:1207.7141, ADS. SPIE 8445.
  • Baron, Monnier & Kloppenburg (2010). A novel image reconstruction software for optical/infrared interferometry. ADS, doi:10.1117/12.857364. SQUEEZE; SPIE 7734.
  • Bekas, Kokiopoulou & Saad (2007). An estimator for the diagonal of a matrix. ADS, doi:10.1016/j.apnum.2007.01.003. Appl. Numer. Math. 57, 1214; the estimator mean(v ⊙ Hv) over Rademacher v for the regularizer Hessian diagonal.
  • Ben-Israel (1966). A Newton-Raphson method for the solution of systems of equations. doi:10.1016/0022-247X(66)90115-6. J. Math. Anal. Appl. 15, 243; Newton with the Moore-Penrose inverse for underdetermined systems; planned minimum-norm set step and C-steps of the orbit-cells search.
  • Benkevitch et al. (2016). Reconstruction of Static Black Hole Images Using Simple Geometric Forms. arXiv:1609.00055. the eccentric ring with a brightness gradient (xs-ring); virgil takes the geometry and derives the visibility by automatic differentiation.
  • Beskos, Pillai, Roberts, Sanz-Serna & Stuart (2013). Optimal tuning of the hybrid Monte Carlo algorithm. doi:10.3150/12-bej414. Bernoulli 19, 1501; acceptance target 0.651.
  • Betancourt (2017). A Conceptual Introduction to Hamiltonian Monte Carlo. arXiv:1701.02434, ADS.
  • Betancourt & Girolami (2013). Hamiltonian Monte Carlo for Hierarchical Models. arXiv:1312.0906, ADS.
  • Bishop (2006). Pattern Recognition and Machine Learning. Springer textbook, section 3.5 (evidence approximation for regularization weights). Not yet verified against arXiv, Crossref or ADS.
  • Blackburn et al. (2020). Closure statistics in interferometric data. arXiv:1910.02062, ADS. ApJ 894, 31 (2020); closure equivalence with gains marginalized, finite gain priors.
  • Blakely et al. (2024). The James Webb Interferometer: Space-based interferometric detections of PDS 70 b and c at 3.8 and 4.8 microns. arXiv:2404.13032, ADS. AJ 169, 137 (2025).
  • Blunt et al. (2017). Orbits for the Impatient: A Bayesian Rejection Sampling Method for Quickly Fitting the Orbits of Long-Period Exoplanets. arXiv:1703.10653, ADS. OFTI.
  • Blunt et al. (2019). orbitize!: A Comprehensive Orbit-fitting Software Package for the High-contrast Imaging Community. arXiv:1910.01756, ADS, GitHub. orbitize!; AJ 159, 89 (2020).
  • Blunt et al. (2024). orbitize! v3: Orbit fitting for the High-contrast Imaging Community. arXiv:2409.11573, ADS, doi:10.21105/joss.06756. orbitize!; JOSS 9, 6756.
  • Born & Wolf (1999). Principles of Optics, 7th edition. doi:10.1017/CBO9781139644181. section 9.2: Zernike radial polynomials, R_n^m(r) ~ r^|m|, the regularity of a smooth field at the origin.
  • Boyd & Vandenberghe (2004). Convex Optimization. doi:10.1017/CBO9780511804441. Cambridge University Press; 9.5.1: the Newton decrement as a stopping rule; planned stop of the orbit-cells set step and C-steps.
  • Brandt et al. (2021). orvara: An Efficient Code to Fit Orbits using Radial Velocity, Absolute, and/or Relative Astrometry. arXiv:2105.11671, ADS, GitHub. orvara; AJ 162, 186.
  • Briggs (1995). High fidelity deconvolution of moderately resolved sources. ADS. PhD thesis, New Mexico Institute of Mining and Technology; the robust weighting; ADS 1995PhDT.......238B. Not yet verified against arXiv, Crossref or ADS.
  • Broderick et al. (2020). THEMIS: A Parameter Estimation Framework for the Event Horizon Telescope. ADS, doi:10.3847/1538-4357/ab91a4. ApJ 897, 139 (2020); Laplace marginalization of station gains.
  • Broucke & Cefola (1972). On the equinoctial orbit elements. ADS, doi:10.1007/BF01228432. Celestial Mechanics 5, 303; Kepler's equation in the eccentricity vector (k, h), regular at e = 0.
  • Buchholz, Chopin & Jacob (2021). Adaptive Tuning of Hamiltonian Monte Carlo Within Sequential Monte Carlo. arXiv:1808.07730, ADS, doi:10.1214/20-ba1222. Bayesian Anal. 16, 745; fixed-L HMC kernels tuned per stage.
  • Byrd, Lu, Nocedal & Zhu (1995). A Limited Memory Algorithm for Bound Constrained Optimization. ADS, doi:10.1137/0916069. SIAM J. Sci. Comput. 16, 1190 (L-BFGS-B); fit(method="lbfgsb") calls scipy's implementation, with the bounds of a linear LinearImage and Uniform priors.
  • Cabezas et al. (2024). BlackJAX: Composable Bayesian inference in JAX. arXiv:2402.10797, GitHub. the adaptive tempered SMC, HMC and pretuning kernels (blackjax >= 1.7.1) that virgil.smc wraps.
  • Candès, Wakin & Boyd (2007). Enhancing Sparsity by Reweighted L1 Minimization. arXiv:0711.1612, ADS. J. Fourier Anal. Appl. 14, 877 (2008).
  • CASA Team (2022). CASAdocs, Synthesis Imaging: Iteration Control (stopping criteria and stop codes). link. Iteration Control, global stopping criteria: divergence 1 is (PeakRes - PrevPeakRes) / PrevPeakRes > 3.0, divergence 2 is (PeakRes - MinPeakRes) / MinPeakRes > 3.0, both over the whole image ignoring the mask (stop codes 5 and 6); minor cycle: (currentPeak - prevPeak) / prevPeak > 0.1, stop code 4 saying 'from minimum'. The page is inconsistent on the minor-cycle reference; virgil uses the running minimum per scale (virgil's choice). Read 2026-10-24. Not yet verified against arXiv, Crossref or ADS.
  • Celeux & Govaert (1992). A classification EM algorithm for clustering and two stochastic versions. doi:10.1016/0167-9473(92)90042-E. Comput. Stat. Data Anal. 14, 315; classification EM: block ascent over labels and parameters; planned inlier-scale step of the orbit-cells search.
  • Chael et al. (2018). Interferometric Imaging Directly with Closure Phases and Closure Amplitudes. arXiv:1803.07088, ADS, doi:10.3847/1538-4357/aab6a8, GitHub. ApJ 857, 23; eht-imaging.
  • Charles et al. (2026). Image reconstruction with the JWST Interferometer. arXiv:2510.10924, ADS. PASA 43, e048 (2026); fit(post_step=): they convolve the log-parameter image with a Gaussian (sigma ~0.25 pix) and renormalise every fitting epoch, to stop flux collapsing into a few pixels.
  • Chavez et al. (2026). Multifrequency Synthesis via CHIBI: Colorful Hierarchical Interferometric Bayesian Imaging. arXiv:2606.04094, ADS.
  • Chen, Feroz & Hobson (2023). Bayesian Posterior Repartitioning for Nested Sampling. arXiv:1908.04655, doi:10.1214/22-BA1323. Bayesian Analysis 18, 695; nested sampling with the prior replaced by any reference and the likelihood by prior x likelihood / reference, same Z; how NS is seeded by the search.
  • Chernoff (1954). On the Distribution of the Likelihood Ratio. ADS, doi:10.1214/aoms/1177728725.
  • Clark (1980). An efficient implementation of the algorithm 'CLEAN'. ADS. A&A 89, 377; bibcode 1980A&A....89..377C; PSF-window minor cycles and the cycle threshold. Not yet verified against arXiv, Crossref or ADS.
  • Cornwell (2008). Multi-Scale CLEAN deconvolution of radio synthesis images. arXiv:0806.2228, ADS, doi:10.1109/jstsp.2008.2006388. IEEE JSTSP 2, 793.
  • Cornwell & Evans (1985). A simple maximum entropy deconvolution algorithm. ADS. A&A 143, 77; the entropy metric as the natural preconditioner in MEM. No DOI or arXiv ID: checked on ADS. Not yet verified against arXiv, Crossref or ADS.
  • Cornwell & Wilkinson (1981). A new method for making maps with unstable radio interferometers. ADS, doi:10.1093/mnras/196.4.1067. MNRAS 196, 1067; self-calibration with a reference antenna fixing the gauge (the reference observation's gain 1; staged fit with the box-edge check).
  • Cotton, Monnier & Baron (2008). 2008 imaging beauty contest. ADS, doi:10.1117/12.788903. Proc. SPIE 7013.
  • Craven & Wahba (1979). Smoothing noisy data with spline functions. doi:10.1007/BF01404567. Numer. Math. 31, 377; the trace of the influence matrix as the effective degrees of freedom of a regularized fit.
  • Currie et al. (2019). No Clear, Direct Evidence for Multiple Protoplanets Orbiting LkCa 15: LkCa 15 bcd are Likely Inner Disk Signals. arXiv:1905.04322, ADS, doi:10.3847/2041-8213/ab1b42. ApJL 877, L3; motivates fitting a disk and a planet together.
  • Czekala et al. (2025). Million Points of Light (MPoL): a PyTorch library for radio interferometric imaging and inference. arXiv:2502.00100, ADS, GitHub, link. software paper; the Gaussian base layer (GaussConvFourier) behind blur_fwhm_mas; ADS bibcode checked.
  • De Cao, Titov & Aziz (2019). Block Neural Autoregressive Flow. arXiv:1904.04676, ADS. BNAF.
  • Del Moral, Doucet & Jasra (2006). Sequential Monte Carlo Samplers. doi:10.1111/j.1467-9868.2006.00553.x. JRSSB 68, 411; SMC samplers and the normalizing-constant estimate, Z as the product of the incremental weights.
  • Desdoigts. zodiax. GitHub. no paper. Not yet verified against arXiv, Crossref or ADS.
  • Desdoigts et al. (2024). Differentiable Optics with dLux I: Deep calibration of Flat Field and Phase Retrieval. arXiv:2406.08703, ADS, GitHub. matrix Fourier transform cross-check; part II is 2406.08704.
  • Desdoigts et al. (2025). AMIGO: a Data-Driven Calibration of the JWST Interferometer. arXiv:2510.09806, ADS. PASA 43, e075 (2026).
  • Dholakia. jax-interferometry. GitHub. the original ELR code (core/ELR.py, core/utils.py), run unmodified at commit 70689ed3dba338d59c98d02e8126a07a3b4e86da by scripts/make_elr_golden.py to produce the golden fixture the port is checked against. Not yet verified against arXiv, Crossref or ADS.
  • Dholakia & Pope (2025). Analytic Interferometry of Rotating Stellar Surfaces. arXiv:2509.25433, ADS, GitHub.
  • Dotter. GDit. GitHub. ELR equations ported via Dholakia from GDit. Not yet verified against arXiv, Crossref or ADS.
  • Drevon et al. (2026). PYRA and MYTHRA: new statistical analysis for image reconstruction. arXiv:2609.15365, ADS. arXiv v1 is 2026.
  • Du & Guyader (2021). Variance estimation in adaptive sequential Monte Carlo. arXiv:1909.13602, doi:10.1214/20-AAP1611. Ann. Appl. Probab. 31, 1021; extends the Eve-index variance estimator to SMC whose tempering schedule is chosen adaptively from the particles.
  • Dumitrescu (2007). Positive Trigonometric Polynomials and Signal Processing Applications. doi:10.1007/978-1-4020-5125-8. Springer (book); the Fejér-Riesz theorem, spectral factorization and the Schur-Levinson recursion behind the positive form of Harmonics. Standard mathematics, credited here and in docstrings, not on the contributors page.
  • Duvert, Young & Hummel (2015). OIFITS 2: the 2nd version of the Data Exchange Standard for Optical (Visible/IR) Interferometry. arXiv:1510.04556, ADS, doi:10.1051/0004-6361/201526405. A&A 597, A8 (2017).
  • Díaz-Cordovés & Giménez (1992). A new nonlinear approximation to the limb-darkening of hot stars. ADS. A&A 259, 227; the square-root limb-darkening law. Not yet verified against arXiv, Crossref or ADS.
  • Edenhofer et al. (2024). Re-Envisioning Numerical Information Field Theory (NIFTy.re): A Library for Gaussian Processes and Variational Inference. arXiv:2402.16683, ADS. JOSS 9, 6593.
  • Efron (1979). Bootstrap Methods: Another Look at the Jackknife. doi:10.1214/aos/1176344552.
  • Elsworth et al. (2025). Measuring the environmental impact of delivering AI at Google Scale. arXiv:2508.15734, ADS. Google's measured 0.24 Wh per median Gemini Apps text prompt; a cross-check for the Gemini figure in the development-carbon estimate (docs/dev_carbon.md).
  • Enßlin (2018). Information theory for fields. arXiv:1804.03350, ADS. Annalen der Physik 531, 1800127 (2019).
  • Enßlin group (Arras, Edenhofer et al.). NIFTy: Numerical Information Field Theory. link. papers: edenhofer2024, arras2021, ensslin2009. Not yet verified against arXiv, Crossref or ADS.
  • Enßlin, Frommert & Kitaura (2008). Information field theory for cosmological perturbation reconstruction and non-linear signal analysis. arXiv:0806.3474, ADS. Phys. Rev. D 80, 105005 (2009).
  • Espinosa Lara & Rieutord (2011). Gravity darkening in rotating stars. arXiv:1109.3038, ADS, doi:10.1051/0004-6361/201117252. A&A 533, A43.
  • Eyer & Bartholdi (1999). Variable stars: Which Nyquist frequency?. arXiv:astro-ph/9808176, doi:10.1051/aas:1999102. A&AS 135, 1; for irregular sampling the usable Nyquist frequency can far exceed that of the mean gap: considered, and not used, for the census's sampling test (scripts/orbit_population.py, nyquist_ok).
  • Fawcett (2006). An introduction to ROC analysis. ADS, doi:10.1016/j.patrec.2005.10.010.
  • Fessler et al. (2005). Toeplitz-based iterative image reconstruction for MRI with correction for magnetic field inhomogeneity. ADS, doi:10.1109/TSP.2005.853152. IEEE Trans. Signal Process. 53, 3393; the Toeplitz embedding of the normal operator AᴴWA on a 2N grid.
  • Fischler & Bolles (1981). Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography. doi:10.1145/358669.358692. Commun. ACM 24, 381; minimal sets of nights as candidate fits (RANSAC).
  • Ford (2006). Improving the Efficiency of Markov Chain Monte Carlo for Analyzing the Orbits of Extrasolar Planets. arXiv:astro-ph/0512634, ADS, doi:10.1086/500802. ApJ 642, 505; the (sqrt(e) cos omega, sqrt(e) sin omega) parameterization, uniform in e; zeta is its half-angle variant.
  • Foreman-Mackey et al. (2021). exoplanet: Gradient-based probabilistic inference for exoplanet data & other astronomical time series. arXiv:2105.01994, ADS, doi:10.21105/joss.03285. JOSS 6, 3285.
  • Fowlie, Handley & Su (2021). Nested sampling with plateaus. arXiv:2010.13884, doi:10.1093/mnras/stab590. MNRAS 503, 1199; likelihood plateaus break nested sampling's compression; the all-outlier floor is one.
  • Frank, Leike & Enßlin (2021). Geometric variational inference. arXiv:2105.10470, ADS. geoVI; not implemented, svi.variational is a different method.
  • Gallenne et al. (2015). Robust high-contrast companion detection from interferometric observations. The CANDID algorithm and an application to six binary stars. arXiv:1505.02715, ADS. CANDID.
  • Gilks & Berzuini (2001). Following a moving target: Monte Carlo inference for dynamic Bayesian models. doi:10.1111/1467-9868.00280. J. R. Stat. Soc. B 63, 127; resample-move: resample weighted draws, then move them with a kernel invariant for the target.
  • Goldfine. frito. GitHub. Not yet verified against arXiv, Crossref or ADS.
  • Goldstein, Poole & Safko (2002). Classical Mechanics (3rd ed.). Addison-Wesley, section 4.5; the Cayley-Klein parameters of a rotation in Euler angles. A textbook with no DOI or ADS record: unverified. Not yet verified against arXiv, Crossref or ADS.
  • Golub & Pereyra (1973). The differentiation of pseudo-inverses and nonlinear least squares problems whose variables separate. doi:10.1137/0710036. SIAM J. Numer. Anal. 10, 413; variable projection: the linear Thiele-Innes constants solved at fixed period and eccentricity.
  • Golub & Welsch (1969). Calculation of Gauss quadrature rules. doi:10.1090/S0025-5718-69-99647-1. Math. Comp. 23, 221; Gauss-Jacobi nodes and weights: the outlier-fraction integral of a degree-K polynomial is exact with ceil((K+1)/2) nodes.
  • Greenbaum et al. (2014). An image-plane algorithm for JWST's non-redundant aperture mask data. arXiv:1411.3446, ADS. ImPlaneIA; ApJ 798, 68 (2015).
  • Greenbaum, Sivaramakrishnan et al.. ImPlaneIA. GitHub. paper: greenbaum2015. Not yet verified against arXiv, Crossref or ADS.
  • Greer. precond.py (coron branch), h-greer/hubble. GitHub. exact entries for small coordinate blocks beside probed large ones, batched one-hot lax.map, imputing non-finite diagonal entries; ideas only, no code copied. Not yet verified against arXiv, Crossref or ADS.
  • Gronau, Sarafoglou, Matzke, Ly, Boehm, Marsman, Leslie, Forster, Wagenmakers & Steingroever (2017). A tutorial on bridge sampling. arXiv:1703.05984, doi:10.1016/j.jmp.2017.09.005. J. Math. Psychol. 81, 80; bridge sampling from MCMC draws to a fitted proposal (compared for the orbit alias-cell evidence).
  • Gross & Vitells (2010). Trial factors for the look elsewhere effect in high energy physics. arXiv:1005.1891, ADS.
  • Guardiani et al. (2022). Causal, Bayesian, & non-parametric modeling of the SARS-CoV-2 viral load distribution vs. patient's age. arXiv:2105.13483, ADS. PLoS ONE 17, e0275011; the Matérn amplitude that NIFTy.re cites (Edenhofer et al. 2024); virgil's MultipoleField is its own construction.
  • Gull (1989). Developments in Maximum Entropy Data Analysis. doi:10.1007/978-94-015-7860-8_4. in Maximum Entropy and Bayesian Methods, Cambridge 1988; the DOI chapter title is the Gull one; Skilling 1989 shares the volume.
  • Hadi & Simonoff (1993). Procedures for the Identification of Multiple Outliers in Linear Models. ADS, doi:10.1080/01621459.1993.10476407. J. Am. Stat. Assoc. 88, 1264; forward addition of observations to the clean set, which the add-one move uses to complete the concentration steps.
  • Han et al. (2020). The extreme colliding-wind system Apep: resolved imagery of the central binary and dust plume in the infrared. arXiv:2008.05834, doi:10.1093/mnras/staa2349. MNRAS (2020); geometric pinwheel model of the Apep plume.
  • Hansen & O'Leary (1993). The Use of the L-Curve in the Regularization of Discrete Ill-Posed Problems. ADS, doi:10.1137/0914086.
  • Hartkopf, McAlister & Franz (1989). Binary star orbits from speckle interferometry. II. Combined visual-spectroscopic orbits of 31 close binaries. ADS, doi:10.1086/115193. AJ 98, 1014; bibcode 1989AJ.....98.1014H.
  • Hattori et al. (2024). jaxoplanet. ADS, doi:10.5281/zenodo.10736936, GitHub. Zenodo; https://github.com/exoplanet-dev/jaxoplanet.
  • Hesterberg (1995). Weighted average importance sampling and defensive mixture distributions. doi:10.1080/00401706.1995.10484303. Technometrics 37, 185; a prior share in the proposal bounds the importance weights by the likelihood maximum over the share.
  • Higham (2002). Accuracy and Stability of Numerical Algorithms. doi:10.1137/1.9780898718027. SIAM, 2nd ed., section 4.2: the rounding error of a sum of n terms is at most about log2(n) eps sum|x_i| for tree-wise summation; the bound used by rounding_dtol where the terms share a sign.
  • Hillen et al. (2016). Imaging the dust sublimation front of a circumbinary disk. arXiv:1603.03023, ADS, doi:10.1051/0004-6361/201628125.
  • Hoffman & Gelman (2011). The No-U-Turn Sampler: Adaptively Setting Path Lengths in Hamiltonian Monte Carlo. arXiv:1111.4246, ADS. NUTS; JMLR 15, 1593 (2014).
  • Hogg, Bovy & Lang (2010). Data analysis recipes: Fitting a model to data. arXiv:1008.4686, ADS.
  • Hogg, Myers & Bovy (2010). Inferring the eccentricity distribution. arXiv:1008.4146, ADS.
  • Honma, Akiyama, Uemura & Ikeda (2014). Super-resolution imaging with radio interferometry using sparse modeling. arXiv:1407.2422, ADS, doi:10.1093/pasj/psu070. PASJ 66, 95.
  • Hutchinson (1989). A stochastic estimator of the trace of the influence matrix for Laplacian smoothing splines. ADS, doi:10.1080/03610918908812806. Commun. Statist. Simul. Comput. 18, 1059; the trace estimator whose Rademacher probes the regularizer part of the diagonal follows (the JᵀJ part uses Gaussian (Jᵀu)² sketches).
  • Högbom (1974). Aperture synthesis with a non-regular distribution of interferometer baselines. ADS. A&AS 15, 417; bibcode 1974A&AS...15..417H. Not yet verified against arXiv, Crossref or ADS.
  • Ireland (2013). Phase Errors in Diffraction-Limited Imaging: Contrast Limits for Sparse Aperture Masking. arXiv:1301.6205, ADS. MNRAS 433, 1718.
  • Ireland, Monnier & Thureau (2006). Monte-Carlo imaging for optical interferometry. arXiv:2007.00716, ADS, doi:10.1117/12.670940. Proc. SPIE 6268 (arXiv posting dated 2020, id checked in ADS); MACIM, whose test of the shortest baselines the large-scale residual check follows.
  • Jasra, Stephens, Doucet & Tsagaris (2011). Inference for Lévy-Driven Stochastic Volatility Models via Adaptive Sequential Monte Carlo. doi:10.1111/j.1467-9469.2010.00723.x. Scand. J. Stat. 38, 1 (online December 2010); ESS-adaptive tempering schedules.
  • Jeffreys (1946). An invariant form for the prior probability in estimation problems. ADS, doi:10.1098/rspa.1946.0056.
  • Jennings et al. (2020). Frankenstein: fitting the visibilities of protoplanetary disc and other radially symmetric sources. arXiv:2005.07709, ADS, GitHub. MNRAS 495, 3209 (2020); github.com/discsim/frank. Ideas only, no code copied (LGPL).
  • Junklewitz et al. (2016). RESOLVE: A new algorithm for aperture synthesis imaging of extended emission in radio astronomy. arXiv:1311.5282, ADS. A&A 586, A76; an IFT imaging algorithm for radio interferometry.
  • Kammerer. fouriever. GitHub. comparison only: compare linear_flux_grid with its lincmap (closure/kernel phases only); no paper found. Not yet verified against arXiv, Crossref or ADS.
  • Kammerer et al. (2020). Increasing the achievable contrast of infrared interferometry with an error correlation model. arXiv:2011.01209, ADS.
  • Kamruddin & Dexter (2013). A geometric crescent model for black hole images. arXiv:1306.3226, doi:10.1093/mnras/stt1068. MNRAS 434, 765; a uniform disk minus an offset uniform disk (the crescent); closed-form visibility.
  • Kass & Raftery (1995). Bayes factors. doi:10.1080/01621459.1995.10476572. JASA 90, 773; the nat scale of a log Bayes factor (below 1: not worth more than a bare mention), the scale against which the SVI loss's remaining fall is capped; the cap is a modelling choice, not Kass & Raftery's.
  • Kennedy (2025). vis-r: a parametric visibility modelling code for radial profiles of discs. arXiv:2502.17619, ADS, GitHub. RASTI (2025); github.com/drgmk/vis-r. Ideas only, no code copied.
  • Kingma & Ba (2014). Adam: A Method for Stochastic Optimization. arXiv:1412.6980, ADS.
  • Kingma et al. (2016). Improving Variational Inference with Inverse Autoregressive Flow. arXiv:1606.04934, ADS. IAF.
  • Kipping (2013). Efficient, uninformative sampling of limb darkening coefficients for two-parameter laws. arXiv:1308.0009, ADS, doi:10.1093/mnras/stt1435.
  • Kluska et al. (2014). SPARCO: a semi-parametric approach for image reconstruction of chromatic objects. arXiv:1403.3343, ADS.
  • Knollmüller & Enßlin (2019). Metric Gaussian Variational Inference. arXiv:1901.11033, ADS. MGVI; laplace_samples shares its covariance form but is a single draw at the MAP, with no iteration.
  • Kraus & Ireland (2012). LkCa 15: A Young Exoplanet Caught at Formation?. arXiv:1110.3808, ADS, doi:10.1088/0004-637X/745/1/5. ApJ 745, 5; motivates fitting a disk and a planet together.
  • Kroupa & Handley (2026). Global structure of the time delay likelihood. arXiv:2602.22307. nested sampling on an aliased (time-delay) likelihood needs many live points to avoid spurious modes; the caution behind running NS as a check, not the default.
  • Kucukelbir et al. (2016). Automatic Differentiation Variational Inference. arXiv:1603.00788, ADS. ADVI; JMLR 18, 14 (2017).
  • Kuramochi et al. (2018). Superresolution Interferometric Imaging with Sparse Modeling Using Total Squared Variation: Application to Imaging the Black Hole Shadow. arXiv:1802.05783, ADS, doi:10.3847/1538-4357/aab6b5. ApJ 858, 56.
  • Kustaanheimo & Stiefel (1965). Perturbation theory of Kepler motion based on spinor regularization. doi:10.1515/crll.1965.218.204. J. reine angew. Math. 218, 204; regularizing maps of Kepler motion.
  • Lachaume (2021). Bias-free model fitting of correlated data in interferometry. arXiv:2104.07082, ADS, doi:10.1017/pasa.2021.20. PASA 38, e033 (2021).
  • Lannelongue, Grealey & Inouye (2021). Green Algorithms: Quantifying the Carbon Footprint of Computation. ADS, doi:10.1002/advs.202100707. Advanced Science 8, 2100707; the method behind the development-carbon estimate (docs/dev_carbon.md).
  • Lawson, Cotton & Hummel (2004). An interferometry imaging beauty contest. ADS, doi:10.1117/12.550710. Proc. SPIE 5491, 886; Eq. 2.
  • Lazareff et al. (2016). Structure of Herbig AeBe disks at the milliarcsecond scale: A statistical survey in the H band using PIONIER-VLTI. arXiv:1611.08428, ADS. A&A 599, A85 (2017).
  • Lee & Greengard (2005). The type 3 nonuniform FFT and its applications. ADS, doi:10.1016/j.jcp.2004.12.004. J. Comput. Phys. 206, 1; a type-1 spread grid evaluated off the integer lattice (the free offsets).
  • Lee & Whiteley (2018). Variance estimation in the particle filter. arXiv:1509.00394, doi:10.1093/biomet/asy028. Biometrika 105, 609; the single-run variance estimator of the normalizing constant from Eve (time-0 ancestor) indices, eq. 4 with phi = 1; needs multinomial resampling.
  • Levi-Civita (1920). Sur la régularisation du problème des trois corps. doi:10.1007/BF02404416. Acta Math. 42, 99; the planar regularizing map.
  • Lindgren, Rue & Lindström (2011). An Explicit Link between Gaussian Fields and Gaussian Markov Random Fields: The Stochastic Partial Differential Equation Approach. doi:10.1111/j.1467-9868.2011.00777.x. JRSS B 73, 423; Matérn fields as solutions of an SPDE.
  • Liu & Nocedal (1989). On the limited memory BFGS method for large scale optimization. ADS, doi:10.1007/BF01589116. L-BFGS.
  • Lucy (2014). Mass estimates for visual binaries with incomplete orbits. arXiv:1309.2868, ADS. A&A 563, A126.
  • Luger et al. (2018). STARRY: Analytic Occultation Light Curves. arXiv:1810.06559, ADS. AJ 157, 64 (2019).
  • Luger, Foreman-Mackey & Hogg (2017). Linear Models for Systematics and Nuisances. arXiv:1710.11136, ADS, doi:10.3847/2515-5172/aa96b5. RNAAS 1, 7.
  • MacKay (1992). Bayesian Interpolation. doi:10.1162/neco.1992.4.3.415. Neural Comput. 4, 415.
  • Malbet et al. (2010). The 2010 Interferometric Imaging Beauty Contest. arXiv:1007.4473, ADS. SPIE 7734.
  • Mantel & Haenszel (1959). Statistical aspects of the analysis of data from retrospective studies of disease. doi:10.1093/jnci/22.4.719. J. Natl. Cancer Inst. 22, 719; stratified test of a common rate across strata; planned orbit population-test assertion.
  • Marquardt (1963). An algorithm for least-squares estimation of nonlinear parameters. doi:10.1137/0111030. J. Soc. Ind. Appl. Math. 11, 431; damping scaled by diag(JᵀJ), per parameter.
  • Martens (2010). Deep learning via Hessian-free optimization. link. ICML 27, 735-742; the Gauss-Newton matrix as the curvature matrix. No DOI or arXiv ID, not on ADS: title and author read from the ICML 2010 PDF. Not yet verified against arXiv, Crossref or ADS.
  • Martinache (2010). Kernel-Phase in Fizeau Inteferometry. arXiv:1009.3933, ADS. ApJ 724, 464 (2010).
  • Martinache (2010). Kernel-Phase in Fizeau Inteferometry. arXiv:1009.3933, ADS, GitHub. xara was formerly pysco; cvis_binary is adapted from it.
  • Mawet et al. (2014). Fundamental limitations of high contrast imaging set by small sample statistics. arXiv:1407.2247, ADS. ApJ 792, 97; the small-sample false-alarm correction, not applied in injection_recovery.
  • McDougall, Pope & Davis (2026). LITMUS: Bayesian lag recovery in reverberation mapping with fast differentiable models. arXiv:2505.09832, doi:10.1017/pasa.2026.10149. PASA 43; hessian_scan: a scan over an aliased variable with optimization and a Laplace evidence per slice, integrated across the grid; prior art for the across-N propagation.
  • Meimon, Mugnier & Le Besnerais (2005). Convex approximation to the likelihood criterion for aperture synthesis imaging. ADS, doi:10.1364/JOSAA.22.002348. JOSA A 22, 2348.
  • Meng & Wong (1996). Simulating ratios of normalizing constants via a simple identity: a theoretical exploration. link. Statistica Sinica 6, 831; bridge sampling and its optimal iterative estimator (compared for the orbit alias-cell evidence). No DOI, arXiv ID or ADS record: title and authors read from the journal page. Not yet verified against arXiv, Crossref or ADS.
  • Mengersen & Tweedie (1996). Rates of convergence of the Hastings and Metropolis algorithms. doi:10.1214/aos/1033066201. Ann. Stat. 24, 101; independence Metropolis-Hastings is uniformly ergodic iff the weight is bounded.
  • Miles (1965). On random rotations in R^3. doi:10.1093/biomet/52.3-4.636. Biometrika 52, 636-639; the isotropic (Haar) distribution of rotations: Omega and omega uniform, cos i uniform. Not on ADS: checked on Crossref.
  • Modrák, Moon, Kim, Bürkner, Huurre, Faltejsková, Gelman & Vehtari (2023). Simulation-based calibration checking for Bayesian computation: the choice of test quantities shapes sensitivity. arXiv:2211.02383. Bayesian Analysis 2023 (arXiv 2022); data-dependent test quantities are valid in SBC; the sky-plane rank of the population test (scripts/orbit_population.py).
  • Monnier et al. (2007). Imaging the Surface of Altair. arXiv:0706.0867, ADS, doi:10.1126/science.1143205. Science 317, 342; the Altair-like truth of notebooks/mwe/mwe_elr_chara.ipynb is inspired by it, not a reproduction.
  • Moran (1975). Quaternions, Haar measure and the estimation of a palaeomagnetic rotation. doi:10.1017/s0021900200047720. J. Appl. Prob. 12, 295-301; Haar measure on SO(3) is the uniform measure on S^3, covered 2:1 by unit quaternions. Not on ADS: checked on Crossref.
  • Morozov (1966). On the solution of functional equations by the method of regularization. Soviet Math. Dokl. 7, 414 (1966); the discrepancy principle. No DOI, and not in ADS or Crossref as of 2026-11-05: unverified. Not yet verified against arXiv, Crossref or ADS.
  • Moré (1978). The Levenberg-Marquardt algorithm: implementation and theory. doi:10.1007/BFb0067700. Lecture Notes in Mathematics 630, 105; per-parameter diagonal scaling makes the step invariant to rescaling the variables.
  • Mérand (2022). Flexible Spectro Interferometric modelling of OIFITS data with PMOIRED. arXiv:2207.11047, ADS, GitHub. PMOIRED.
  • Mérand et al.. CANDID: Companion Analysis and Non-Detection in Interferometric Data. GitHub. algorithm paper: gallenne2015. Not yet verified against arXiv, Crossref or ADS.
  • Neal (2011). MCMC using Hamiltonian dynamics. arXiv:1206.1901. Handbook of Markov Chain Monte Carlo, ch. 5; randomized step size and trajectory length for HMC.
  • Nocedal & Wright (2006). Numerical Optimization. doi:10.1007/978-0-387-40065-5. Springer, 2nd ed.; ch. 12 and 16.7: KKT and projected-gradient conditions (kkt_report), KKT sign condition of a bound (fit's bound_stall); ch. 3-4 and 10.3: safeguarded steps in a bracket, trust-region predicted reduction, Newton and Levenberg-Marquardt in curved valleys; planned set step of the orbit-cells search.
  • Offringa et al. (2014). WSCLEAN: an implementation of a fast, generic wide-field imager for radio astronomy. arXiv:1407.1943, ADS. MNRAS 444, 606; bibcode 2014MNRAS.444..606O; the WSClean argument conventions radio_clean follows.
  • Oudoumanessah & Iollo. nufftax. GitHub. no paper; its spreader and kernel parameters (private API, nufftax >= 0.6.1). Not yet verified against arXiv, Crossref or ADS.
  • Papaspiliopoulos, Roberts & Sköld (2007). A General Framework for the Parametrization of Hierarchical Models. ADS, doi:10.1214/088342307000000014. Statistical Science 22, 59; centred and non-centred parameterizations.
  • Pauls et al. (2005). A Data Exchange Standard for Optical (Visible/IR) Interferometry. arXiv:astro-ph/0508185, ADS, doi:10.1086/444523. OIFITS 1; PASP 117, 1255.
  • Pawitan (2001). In All Likelihood: Statistical Modelling and Inference Using Likelihood. doi:10.1093/oso/9780198507659.001.0001. Oxford University Press (book); the profile likelihood of the noise scale, sigma_hat = sqrt(chi2 / 2N). Not on ADS: checked on Crossref.
  • Phipson & Smyth (2010). Permutation P-values Should Never Be Zero: Calculating Exact P-values When Permutations Are Randomly Drawn. doi:10.2202/1544-6115.1585. Stat. Appl. Genet. Mol. Biol. 9, 39; (1 + #)/(1 + n) permutation p-value.
  • Pope et al.. jaxbessel. GitHub. CEPHES (Moshier, https://www.netlib.org/cephes/) translation, adapted from harmonix; the Cephes credit lives here and in the jaxbessel repository. Not yet verified against arXiv, Crossref or ADS.
  • Price-Whelan et al. (2017). The Joker: A custom Monte Carlo sampler for binary-star and exoplanet radial velocity data. arXiv:1610.07602, ADS. ApJ 837, 20.
  • Price-Whelan et al. (2020). Close Binary Companions to APOGEE DR16 Stars: 20,000 Binary-star Systems Across the Color-Magnitude Diagram. arXiv:2002.00014, ADS. ApJ 895, 2; The Joker at scale.
  • Quirrenbach et al. (1996). Angular diameter and limb darkening of Arcturus. ADS. A&A 312, 160; bibcode 1996A&A...312..160Q. Not yet verified against arXiv, Crossref or ADS.
  • Rasmussen & Williams (2006). Gaussian Processes for Machine Learning. doi:10.7551/mitpress/3206.001.0001, link. MIT Press (book); GP terminology, Matérn kernels, marginal likelihood; A.9 matrix inversion and determinant lemmas (arrow_at).
  • Rau & Cornwell (2011). A multi-scale multi-frequency deconvolution algorithm for synthesis imaging in radio interferometry. arXiv:1106.2745, ADS, doi:10.1051/0004-6361/201117104. A&A 532, A71.
  • Ritter & Tanner (1992). Facilitating the Gibbs Sampler: The Gibbs Stopper and the Griddy-Gibbs Sampler. doi:10.1080/01621459.1992.10475289. JASA 87, 861; griddy Gibbs: draw a scalar conditional by tabulating it on a grid (a review toy for the orbit alias-cell search).
  • Roettenbacher et al. (2016). No Sun-like dynamo on the active star ζ Andromedae from starspot asymmetry. arXiv:1709.10107, ADS, doi:10.1038/nature17444. Nature 533, 217; CHARA/MIRC starspot imaging that motivates HarmonixModel.
  • Rousseeuw (1984). Least median of squares regression. doi:10.1080/01621459.1984.10477105. J. Am. Stat. Assoc. 79, 871; elemental subsets as starts of a robust fit.
  • Rousseeuw & Van Driessen (1999). A fast algorithm for the minimum covariance determinant estimator. doi:10.1080/00401706.1999.10485670. Technometrics 41, 212; concentration steps (C-steps): relabel inliers, refit, until the labels repeat.
  • Rudin, Osher & Fatemi (1992). Nonlinear total variation based noise removal algorithms. ADS, doi:10.1016/0167-2789(92)90242-F. Physica D 60, 259.
  • Rue & Held (2005). Gaussian Markov Random Fields: Theory and Applications. doi:10.1201/9780203492024. Chapman & Hall/CRC (book).
  • Rue, Martino & Chopin (2009). Approximate Bayesian inference for latent Gaussian models by using integrated nested Laplace approximations. doi:10.1111/j.1467-9868.2008.00700.x. J. R. Stat. Soc. B 71, 319; INLA: quadrature in a low-dimensional parameter with conditional Laplace; planned ridge proposal component in rho for orbit cells.
  • Ruffio et al. (2018). A Bayesian Framework for Exoplanet Direct Detection and Non-Detection. arXiv:1809.08261, ADS.
  • Sallum et al. (2015). Accreting protoplanets in the LkCa 15 transition disk. arXiv:1511.07456, ADS, doi:10.1038/nature15761. Nature 527, 342; motivates fitting a disk and a planet together.
  • Schraudolph (2002). Fast Curvature Matrix-Vector Products for Second-Order Gradient Descent. doi:10.1162/08997660260028683. Neural Comput. 14, 1723; Gauss-Newton matrix-vector products by forward-mode differentiation. Not on ADS: checked on Crossref.
  • Schwab (1984). Relaxing the isoplanatism assumption in self-calibration; applications to low-frequency radio interferometry. ADS. AJ 89, 1076; bibcode 1984AJ.....89.1076S; the Cotton-Schwab major cycle (residual from the visibilities). Cotton is cited there as a private communication. Not yet verified against arXiv, Crossref or ADS.
  • Self & Liang (1987). Asymptotic properties of maximum likelihood estimators and likelihood ratio tests under nonstandard conditions. ADS, doi:10.1080/01621459.1987.10478472. Nonstandard conditions: a parameter on the boundary, or a nuisance parameter unidentified under the null.
  • Selig et al. (2013). NIFTY - Numerical Information Field Theory - a versatile Python library for signal inference. arXiv:1301.4499, ADS.
  • Shoemake (1992). Uniform Random Rotations. doi:10.1016/b978-0-08-050755-2.50036-1. Graphics Gems III, 124; unit quaternions and the Haar measure of SO(3).
  • Sivaramakrishnan et al. (2022). The Near Infrared Imager and Slitless Spectrograph for the James Webb Space Telescope. IV. Aperture Masking Interferometry. arXiv:2210.17434, ADS. PASP 135, 015003 (2023).
  • Skilling (1989). Classic Maximum Entropy. doi:10.1007/978-94-015-7860-8_3. Maximum Entropy and Bayesian Methods, Cambridge 1988; the un-normalised entropy of MaxEntropy on a LinearImage. docs/contributors.md links its Skilling 1989 entry to DOI ..._4, which Crossref gives as the Gull 1989 chapter.
  • Skilling (2006). Nested sampling for general Bayesian computation. doi:10.1214/06-BA127. Bayesian Analysis 1, 833; nested sampling, candidate (d) for the per-cell evidence; robust to first-order phase transitions where tempering fails.
  • Skilling & Bryan (1984). Maximum entropy image reconstruction: general algorithm. ADS. MNRAS 211, 111; bibcode 1984MNRAS.211..111S; precondition="diag" generalizes its entropy metric to the local curvature of data and regularizers. Not yet verified against arXiv, Crossref or ADS.
  • Soulain et al.. AMICAL: aperture masking interferometry calibration and analysis. GitHub. separate package, bibliography only; paper reference to add (SPIE 2020). Not yet verified against arXiv, Crossref or ADS.
  • Soummer et al. (2007). Fast computation of Lyot-style coronagraph propagation. arXiv:0711.0368, ADS. matrix Fourier transform.
  • Starck, Murtagh & Fadili (2010). Sparse Image and Signal Processing: Wavelets, Curvelets, Morphological Diversity. book (CUP). Not yet verified against arXiv, Crossref or ADS.
  • Strang (1999). The Discrete Cosine Transform. ADS, doi:10.1137/S0036144598336745. SIAM Review 41, 135; the DCT-II vectors are the eigenvectors of the second-difference matrix with symmetric (Neumann) boundaries.
  • Surjanovic, Biron-Lattes, Tiede et al. (2023). Pigeons.jl: Distributed Sampling From Intractable Distributions. arXiv:2308.09769, ADS. parallel tempering prior art next to Octofitter.
  • Talts, Betancourt, Simpson, Vehtari & Gelman (2018). Validating Bayesian inference algorithms with simulation-based calibration. arXiv:1804.06788. rank-uniformity check of a posterior against its own prior draws; the population test of the cell fit (scripts/orbit_population.py).
  • Thiele (1883). Neue Methode zur Berechung von Doppelsternbahnen. ADS. Astron. Nachr. 104, 245; title as spelled by ADS. Not yet verified against arXiv, Crossref or ADS.
  • Thiébaut & Young (2017). Principles of image reconstruction in optical interferometry: tutorial. arXiv:1708.08390, ADS. JOSA A 34, 904.
  • Thompson et al. (2023). Octofitter: Fast, Flexible, and Accurate Orbit Modelling to Detect Exoplanets. arXiv:2402.01971, ADS, doi:10.3847/1538-3881/acf5cc, GitHub. AJ 166, 164 (2023); the arXiv preprint was posted in February 2024, after publication.
  • Tiede et al. (2025). Hierarchical Interferometric Bayesian Imaging. arXiv:2511.17706, ADS. HIBI/Comrade.
  • Tierney & Kadane (1986). Accurate Approximations for Posterior Moments and Marginal Densities. doi:10.1080/01621459.1986.10478240. JASA 81, 82; Laplace's method for marginal likelihoods (the baseline for the orbit alias-cell evidence).
  • Tuthill et al. (2008). The prototype colliding-wind pinwheel WR 104. arXiv:0712.2111, doi:10.1086/527286. ApJ 675, 698; uniformly rotating Archimedean spiral fitted to the WR 104 plume, with the measured brightness profile along it.
  • Tuthill, Monnier & Danchi (1999). A dusty pinwheel nebula around the massive star WR 104. arXiv:astro-ph/9904092, doi:10.1038/19033. Nature 398, 487; the dust plume follows a linear (Archimedean) spiral; geometry only.
  • van Hamme (1993). New Limb-Darkening Coefficients for Modeling Binary Star Light Curves. ADS, doi:10.1086/116788. AJ 106, 2096; the quadratic law the square-root law is compared with.
  • VanderPlas (2017). Understanding the Lomb-Scargle Periodogram. arXiv:1703.09824, ADS. ApJS 236, 16 (2018).
  • Veach & Guibas (1995). Optimally combining sampling techniques for Monte Carlo rendering. doi:10.1145/218380.218498. SIGGRAPH '95, 419; the balance heuristic, the partition of unity that splits the bridge per mode.
  • Vehtari et al. (2015). Pareto Smoothed Importance Sampling. arXiv:1507.02646, ADS. JMLR 25, 72 (2024).
  • Vehtari et al. (2021). Rank-normalization, folding, and localization: An improved R-hat for assessing convergence of MCMC. arXiv:1903.08008, ADS. Bayesian Analysis 16, 667; R-hat is trusted only with bulk ESS above 100 per chain.
  • virgil (own design) (2026). MultipoleField: a multipole-expanded Gaussian-process image on a polar disc-frame grid. No published source for the construction: polar grid, radial Matérn Cholesky, tanh taper, multipoles and edge fade are virgil's own; it shares only the exponential-of-a-GP image and whitened latents with the IFT papers. Not yet verified against arXiv, Crossref or ADS.
  • virgil (own design) (2026). Tempering an evidence for correlated spectral channels. No published source. The effective-channel factor from neighbouring-channel residual covariance is virgil's own method; provisional, and no test checks that tempered evidences are calibrated. Not yet verified against arXiv, Crossref or ADS.
  • virgil (own design) (2026). Short-arc motion constraints for the orbit period search. No published source. Provisional; not independently validated. Not yet verified against arXiv, Crossref or ADS.
  • virgil (own design) (2026). Period-alias band search for sparsely sampled orbits. No published source. The search over cycle-count bands, its nats thresholds and its night-outlier mixture (marginalized rate, Beta(1/2, 1/2) (Jeffreys)) are virgil's own design, provisional and not independently validated end to end. Not yet verified against arXiv, Crossref or ADS.
  • Wahba (1983). Bayesian "Confidence Intervals" for the Cross-Validated Smoothing Spline. doi:10.1111/j.2517-6161.1983.tb01239.x. J. R. Stat. Soc. B 45, 133; the effective degrees of freedom Tr[K (K + H_R)^-1] of a regularized fit (cf. Hansen 1998, Rank-Deficient and Discrete Ill-Posed Problems, SIAM).
  • Wilks (1938). The Large-Sample Distribution of the Likelihood Ratio for Testing Composite Hypotheses. doi:10.1214/aoms/1177732360.
  • Wright & Howard (2009). Efficient fitting of multiplanet Keplerian models to radial velocity and astrometry data. arXiv:0904.3725, ADS. ApJS 182, 205.
  • Yallup, Kroupa & Handley (2026). Nested Slice Sampling: Vectorized Nested Sampling for GPU-Accelerated Inference. arXiv:2601.23252. the BlackJAX nested slice sampler (blackjax.nss, >= 1.7) measured as candidate (d) on the leg-1 toys.
  • Yoshida et al. (2026). Direct Retrieval of Protoplanetary Disk Dust Properties Using Auto-differentiable Gaussian Processes and Their Application to the HD 169142 Disk. arXiv:2607.12618, ADS, doi:10.3847/1538-4357/ae89ad, GitHub. ApJ 1007, 198 (2026); github.com/tomyoshida/frap. Ideas only, no code copied: whitened Gaussian-process latents (g = mu + L z) at radial nodes with a dense kernel, sampled by NUTS, and a radial GP per physical quantity.
  • Zawadzki et al. (2023). Regularized Maximum Likelihood Image Synthesis and Validation for ALMA Continuum Observations of Protoplanetary Disks. arXiv:2209.11813, ADS, doi:10.1088/1538-3873/acdf84. PASP 135, 064503 (MPoL); the RML objective for ALMA visibilities that virgil.radio reproduces (design register A).
  • Zhou, Johansen & Aston (2016). Toward Automatic Model Comparison: An Adaptive Sequential Monte Carlo Approach. doi:10.1080/10618600.2015.1060885. JCGS 25, 701; conditional ESS for the next temperature.
  • Zhu, Byrd, Lu & Nocedal (1997). Algorithm 778: L-BFGS-B: Fortran subroutines for large-scale bound-constrained optimization. doi:10.1145/279232.279236. ACM Trans. Math. Softw. 23, 550; the L-BFGS-B code scipy wraps. Not yet verified against arXiv, Crossref or ADS.