About the project

AstroJyotish.info is a free reference on Vedic astrology with a natal chart calculator. The whole site — the texts on the planets, the houses, the signs, the nakshatras and the yogas, and the calculation itself — is built around a single reading standard: the «Surya-Siddhanta» settings of Vasily Tushkin's school of Vedic astrology, the same ones used by the Jagannatha Hora program.

Texts collected from open sources are combined with the calculated results, so that a chart can be read without leafing through a shelf of books: next to each placement stands the passage that explains it.

Why the calculation had to be rebuilt

Jagannatha Hora is a desktop program for Windows. It cannot be embedded in a website, it has no interface for automated queries, and the «Surya-Siddhanta» mode that this project relies on is one of its internal settings rather than a documented format. For a long time the site worked around this the only way it could: charts were exported from the program by hand and the text export was parsed on the server.

That approach has an obvious ceiling. Every new chart depends on someone launching a Windows program; the queue grows; a machine that goes down stops the site. So the calculation part was reverse-engineered and written again from scratch in modern programming languages — Python for the research and the ephemeris, PHP for the site itself.

How the reverse engineering was done

The reference was a corpus of 2413 charts previously produced by the program and parsed into structured data. Everything in the new implementation is checked against that corpus, and the check is honest: the parameters were fitted on 60% of the charts and the figures are reported on the remaining 40%, which the optimiser never saw.

The first thing to fall was the ayanamsa. It turned out to be a strictly linear function of time — 54 arc-seconds per year, with its zero near the year 499 — and not one of the roughly forty-five ayanamsa modes of the Swiss Ephemeris reproduces it: the closest of them is off by 77 arc-seconds. The program computes that formula itself.

The planetary longitudes proved to be an entirely different matter. No sidereal mode of a modern ephemeris fits them: the error never falls below about an arc-degree, and, crucially, it differs from planet to planet. That ruled out a constant shift and pointed at a different planetary theory altogether — the classical one of the Surya-Siddhanta, with mean motions, a manda equation for the apsis, a shighra equation for conjunction, and a variable epicycle interpolated across the quadrants.

Reconstructing that theory took three findings. Each planet needed a small shift of its own mean longitude — a bija correction to the mean motions, which the fit recovered on its own; the fitted shighra and mandochcha values then landed almost exactly on the canonical numbers of the treatise, which is what confirmed the shape of the model. Next, the canonical rule of applying half the manda equation at the second step turned out to be applied in full in practice: given freedom, the coefficient converged on 1.0 for every planet and improved Jupiter from 2.43' to 0.12'. And last, the residual of the Moon decomposed into an annual and a semi-annual wave — precisely the two components of the equation of time, bhujantara and udayantara, whose fitted coefficients came out at −0.002800 and +0.002726 against the theoretical −1/360 and +1/360.

What the result is

All nine bodies are now reproduced to better than half an arc-minute of median error, where a modern ephemeris with a shifted ayanamsa is out by tens or hundreds of arc-minutes: Rahu and Ketu to 0.00', the Sun and Jupiter to 0.04', Saturn to 0.15', the Moon to 0.21', Mercury to 0.23', Venus to 0.25' and Mars to 0.44'. That is well inside the precision at which a chart reading changes nothing.

What this means in practice is simple: the chart on this site is calculated on the server, immediately, without a queue and without a Windows machine in the loop — and it agrees with what Jagannatha Hora would have produced in the same mode. The divisional charts, the dashas, the ashtakavarga and the avasthas are reproduced by the same route, each verified against the same corpus.

Honesty about the limits

The site is a reference, not an oracle. The calculation reproduces a particular tradition of computation, not astronomical reality: the Surya-Siddhanta is a medieval model, and its longitudes deliberately differ from those of modern ephemerides. This is the point — a reading built on one school's numbers should be done with that school's numbers.

The interpretive texts come from open sources and from classical treatises; they are compiled, not invented, and they contradict one another as often as classical sources normally do. Treat them as material for your own thinking rather than as a verdict. And nothing on this site replaces a doctor, a lawyer or a therapist.

The project carries no advertising and does not sell visitors' data. It is kept up on voluntary contributions — see the consultation page. Questions, corrections and reports of errors in the texts or in the calculation are welcome at info@astrojyotish.info.