A reproducible calculation
Eros, checked.
Asteroid 433. A pinned ephemeris. An independent comparison.
What the test establishes
The retained Swiss Ephemeris 2.10.03 engine evaluates Eros at the requested birth instant. Six selected epochs were compared with NASA JPL Horizons, using a geocentric observer (500@399), apparent ecliptic longitude and latitude of date (quantity 31), no atmospheric refraction, and matching times. Horizons uses the IAU76/80 ecliptic frame; the small residual includes differences in models and orbital solutions.
All tested longitude and latitude differences are below one arcsecond. The largest is 0.521218 arcseconds. These sampled comparisons do not independently validate every possible birth date.
| UTC / UT test epoch | Longitude difference, arcsec | Latitude difference, arcsec |
|---|---|---|
| 1800-06-15 12:00 | 0.520366 | 0.025751 |
| 1950-01-01 00:00 | 0.090113 | 0.034426 |
| 1988-11-12 14:30 | 0.033589 | 0.012654 |
| 2000-01-01 12:00 | 0.046298 | 0.018050 |
| 2026-10-21 12:00 | 0.047171 | 0.025561 |
| 2095-06-15 12:00 | 0.521218 | 0.056711 |
Before 1972 the inherited Swiss time convention treats the input as UT1; modern dates use UTC. The interface resolves historical civil time using bundled IANA time-zone data.
The fictional example
Alex Rowan: 12 November 1988 at 14:30, London, 51.5074° N / 0.1278° W. Europe/London is GMT (UTC+0). Eros: 12°52′ Aries, retrograde; Porphyry house 1, Whole Sign house 2. Eros quincunx Pluto: 0.03256097°, or 0°01′57″ rounded.
The ten planets were also checked against Horizons at that instant; the largest longitude/latitude difference is 0.085858 arcseconds. The true lunar node is not part of this independent planetary comparison.
Porphyry cusps were checked against a separate quadrant-trisection formula; Whole Sign cusps against 30° sign boundaries. The separate formula shares Ascendant and Midheaven inputs with Swiss Ephemeris. House assignments use ecliptic longitude, including Eros. A latitude-sensitive mundane house method would be a different convention.
Browser parity and input safeguards
13 test scenarios and 7,007 numerical comparisons agree between the native engine and the browser build. The greatest absolute numerical difference was 2.11e-08. Tests include the example, selected historical/future dates, a rising-sign boundary, a nonexistent daylight-saving time, an ambiguous repeated hour and both explicit occurrences. Unknown birth time is rejected for this house-based demo.
Eros speed was checked by a central finite difference. The asteroid file is hash-checked before calculation; there is no silent substitute for missing data. The browser verifies the entire bundled data manifest.
The near-exact Eros–Pluto quincunx is included under a proposed 1° rule. Major Eros aspects use 2°. These are editorial conventions for Sabrina’s approval. The calculation check does not validate the symbolic interpretation.
Sources and reproducibility
NASA JPL Horizons documentation
Astrodienst distribution documentation
Eros ephemeris file from the documented distribution mirror
SHA-256: 887549f513f0e88bfb64128b4be48c0a3f6697d9292f1931104251956c985e9b
Prototype source, verification summaries and licences · New six-page Eros review