Calculation checks

21 October 2026 · Independent proposal for Timothy Halloran. The prototype does not claim his endorsement or reproduce a licensed teaching system.

Shared calculation engine

Swiss Ephemeris 2.10.03 with local DE441-based Swiss files; geocentric, tropical, apparent positions, Placidus houses and true lunar nodes. Historical time zones use IANA data. Calculations run in the browser. The demo refuses unavailable Placidus houses, missing birth times and unresolved ambiguous local times. Supported dates: 1800–2095.

TH-PATTERN-1.1: conjunction/opposition 6°, square/trine 5°, sextile 3°. Node contacts use a 3° orb. Modern sign rulers are an explicit prototype choice requiring Timothy’s confirmation. The Pluto polarity point is exactly opposite Pluto; the South Node is exactly opposite the North Node. Each axis is counted once, and neither an axis nor the polarity point is an extra physical planet.

Fictional reference case

Maya Ellis · 16 May 1989 · 08:40 Bristol · 51.4545 N, 2.5879 W · Europe/London. Historical BST gives 07:40 UTC. Pluto is 13°27′51″ Scorpio in house 5, polarity 13°27′51″ Taurus in house 11. The North Node is 1°09′00″ Pisces in house 9; the South Node is in Virgo, house 3.

Independent reference and tests

• The independent reference was generated on 12 October using Skyfield 1.55 and JPL DE440s. On 21 October the current engine was compared with that retained reference. This is not a newly generated independent ephemeris run.
• Ten planetary longitude differences are below 0.001 arcsecond. The true-node difference from an independently constructed lunar orbital plane is 0.00644 arcsecond.
• Independent Placidus semi-arc geometry differs by at most 0.319 arcsecond at a house cusp.
• All 13 printed positions/houses and the 11 selected aspect orbs agree with the unchanged report at its displayed precision.
• 121 full recalculations, every five seconds from 08:35 to 08:45, preserve all 13 house placements and the qualifying aspect membership. This is a finite sensitivity test for this case.
• 14 native-versus-browser scenarios compare 3,561 numeric values: historical and future dates, daylight-saving gaps and repeated times, missing time and polar-house refusal. Maximum numerical difference: 0.0000000212. These test cases do not establish every possible input.

The on-screen house sensitivity table shifts the house cusps while holding the original planet positions fixed; that limitation is labelled there. The separate 121-case test recalculates everything.

Interpretation and report

The short guide uses actual sign, house and selected aspect data to frame original questions. It is not the complete 28-page personalised report. Only text supplied in the four optional fields describes the visitor’s experience. No personal history or another person’s motives are inferred. The thirty days organise observations, not dated transits or predictions.

The fixed fictional PDF is the unchanged partner-review edition from 12 October. Its original Moshier calculation background is preserved; the current Swiss-file engine agrees at the printed precision. All 28 pages were individually visually reviewed on 12 October, and the file’s identity was verified on 21 October. No new PDF has been generated from the visitor’s inputs.

Original fictional sample · Source, tests documentation and licences · Swiss Ephemeris · Skyfield · JPL ephemerides

Before a live service

Agree the method, interpretation library, editorial process, brand permission, prices and commercial arrangement. Connect the approved website, consent and order records, test checkout, verified payment events, production checks and private delivery. None of those live integrations is connected here. Production hosting and the Swiss Ephemeris licence model require agreement. Source is provided under AGPL-3.0-or-later.