Calculations used to generate a chart

[Guest guest]

Hello,

 

I have just begun my study of Vedic astrology. I have purchased the book,

"Beneath a Vedic Sky" which arrived complete with a Limited GJ

program. As a programmer myself, I feel the need to complement Das on his

fine work. I can only imagine the enormous amount of study and

programming man-hours that went into its creation.

 

However, to facilitate my understanding of Jyotish, I was hopeful that

someone on this list could point me to a reference (book, url, etc.) that

details the mathmatical calculations used in creating a natal chart. Any

pointer will be of great help.

 

I am able to calculate a Western Tropical based chart 'by hand', but would

like to be able to do the same with a Vedic/sidereal chart. I've yet to

find such a reference on the net, although surely there is one.

 

Thank you for you time and consideration of this request.

 

 

-Scott

[Guest guest]

Hey Scott--

 

You know, there just aren't that many resources out there which get into

the calculation details. As a developer myself, I can tell you that it IS

out there, just don't expect to find it spoon-fed to you. FWIW, I used the

same approach you want to use--rather than just using the software and

taking its output as gospel, I went through the calculations myself to make

sure I understood what GJ was doing. For the record, I still consider

myself a rank amateur at this, but perhaps you can benefit from some of my

experience.

 

B.V. Raman's "110 Year Ephemeris of Planetary Positions" contains pretty

detailed instructions on how to erect a horoscope with a sidereal ephemeris.

 

His book "A Manual of Hindu Astrology" is excellent, but remember that

he wrote this book in 1935, and writing styles have changed quite a bit

since then. Even so, for my own development work, it's been quite

satisfactory with plenty of worked examples. You'll want this book for

details on how to compute the vargas (divisional, or harmonic) charts. Some

of the vargas are rather tedious until you become accustomed to the writing

style.

 

If you're looking to develop, you have two options: You can either use

high-precision ephemerides (the subject of a brief discussion earlier this

week), or you can use harmonic models. I suggest the (free for

non-commercial development) Swiss Ephemeris as an excellent high-precision

electronic reference; the distribution includes useful code examples.

 

For a harmonic model which can be computed without the ephemeris (but

with corresponding loss of accuracy, which may or may not be an issue

depending upon your application), check out Paul Schlyter's online paper

"Computing Planetary Positions--A Tutorial with Worked Examples", with code

examples in Basic, Pascal, and Fortran. Remember to convert between

equatorial and ecliptic coordinates. This is an astronomical work, not an

astrological one. I don't have the URL at hand, but I'm sure I can dig it

up if you can't find it.

 

Other astronomical algorithms are presented in Jean Meeus' "Astronomical

Algorithms" and his "Astronomical Formulae for Calculators". The two works

use different reference points for right ascension (that's something else

you have to keep track of, btw). The calculator book uses the B1950 epoch,

which uses the equinox and mean equator of December 31, 1949 at 2209Z. The

other, and currently prevailing model, uses the equinox and mean equator of

January 1, 2000 at 1200Z.

 

Meeus' chapters on interpolation are especially useful if you are

analyzing tabular data.

 

Shad Bala is pretty intricate, too; B.V. Raman's "Graha and Bhava Balas"

walks through that. I'm still coding that. Das, where on earth did you

find the time?

 

For calculating the ascendant, you may have another job on your hands,

but not as bad. Alan Leo's book "Casting the Horoscope" contains some

useful formulae, but this was written in the 1800s, so much of the work

involves how to use log and trig tables. For quick calculations, of course,

better to use an ascendant table. (The Swiss Ephemeris can calculate that,

too.)

 

Penultimately, the ayanamsa. I haven't had much luck using any linear

interpolation or extrapolation (Mani is already getting red in the face, I

know) to estimate it. My linearly-interpolated results rarely coincide what

what GJ computes, and I'm not doubting GJ's accuracy here--either my

reference data is ratty or I'm using too simplistic of an interpolation

model; one day, I'll try a Lagrange spline. So, instead, I use the Swiss

Ephemeris to compute the coordinates of Spica (the first point of Libra) and

add 180 degs (plus like an arc minute or something) to it. This gives the

Lahiri first point of Aries. Emotionally, that approach is more satisfying,

anyway. If somebody knows what I'm doing wrong, please tell.

 

And finally, be suspicious about algorithms you find on the web. I

found one article, supposedly written by an ASTRONOMER no less, who claimed

you could determine the lunar naksatra by computing the modulus of [ the

date in question - some reference date, in days ] by 27.3 (a very rough

approximation of the lunar sidereal month). The moon's motion just isn't

that simple that it can be modeled accurately with only three sig figs. So,

if it's simple, suspect it: The only thing simple is getting it wrong.

Have fun, though ... it's a worthy occupation of time.

 

jpd

[Guest guest]

Hi Phil,

 

Thank you so much for the information. As a rule of thumb I have learned

not to expect much in this life spoon fed to me, but I did think the

information would be more readily available on the web than it appears to be

(perhaps I'm just not structuring my search strings correctly).

 

However, your caution to be skeptical of algorithms found on the web is most

appreciated. I often forget to do that at the beginning of a project, only

to find out how careless I was at an inopportune time.

 

Also, thank you for citing the various texts you have found helpful. You

have definitely given me a great deal to start with.

 

Indeed, Das, where *did* you find the time. The more I play with GJ, the

more I want to applaud your skills and effort. GJ is a gorgeous program.

 

Namaste,

 

-scott

 

-

"Phil Draughon" <jpd

<gjlist>

Friday, March 30, 2001 4:05 PM

Re: [gjlist] Calculations used to generate a chart

 

 

> Hey Scott--

>

> You know, there just aren't that many resources out there which get

into

> the calculation details. As a developer myself, I can tell you that it IS

> out there, just don't expect to find it spoon-fed to you. FWIW, I used

the

> same approach you want to use--rather than just using the software and

> taking its output as gospel, I went through the calculations myself to

make

> sure I understood what GJ was doing. For the record, I still consider

> myself a rank amateur at this, but perhaps you can benefit from some of my

> experience.

>

> B.V. Raman's "110 Year Ephemeris of Planetary Positions" contains

pretty

> detailed instructions on how to erect a horoscope with a sidereal

ephemeris.

>

> His book "A Manual of Hindu Astrology" is excellent, but remember that

> he wrote this book in 1935, and writing styles have changed quite a bit

> since then. Even so, for my own development work, it's been quite

> satisfactory with plenty of worked examples. You'll want this book for

> details on how to compute the vargas (divisional, or harmonic) charts.

Some

> of the vargas are rather tedious until you become accustomed to the

writing

> style.

>

> If you're looking to develop, you have two options: You can either

use

> high-precision ephemerides (the subject of a brief discussion earlier this

> week), or you can use harmonic models. I suggest the (free for

> non-commercial development) Swiss Ephemeris as an excellent high-precision

> electronic reference; the distribution includes useful code examples.

>

> For a harmonic model which can be computed without the ephemeris (but

> with corresponding loss of accuracy, which may or may not be an issue

> depending upon your application), check out Paul Schlyter's online paper

> "Computing Planetary Positions--A Tutorial with Worked Examples", with

code

> examples in Basic, Pascal, and Fortran. Remember to convert between

> equatorial and ecliptic coordinates. This is an astronomical work, not an

> astrological one. I don't have the URL at hand, but I'm sure I can dig it

> up if you can't find it.

>

> Other astronomical algorithms are presented in Jean Meeus'

"Astronomical

> Algorithms" and his "Astronomical Formulae for Calculators". The two

works

> use different reference points for right ascension (that's something else

> you have to keep track of, btw). The calculator book uses the B1950

epoch,

> which uses the equinox and mean equator of December 31, 1949 at 2209Z.

The

> other, and currently prevailing model, uses the equinox and mean equator

of

> January 1, 2000 at 1200Z.

>

> Meeus' chapters on interpolation are especially useful if you are

> analyzing tabular data.

>

> Shad Bala is pretty intricate, too; B.V. Raman's "Graha and Bhava

Balas"

> walks through that. I'm still coding that. Das, where on earth did you

> find the time?

>

> For calculating the ascendant, you may have another job on your hands,

> but not as bad. Alan Leo's book "Casting the Horoscope" contains some

> useful formulae, but this was written in the 1800s, so much of the work

> involves how to use log and trig tables. For quick calculations, of

course,

> better to use an ascendant table. (The Swiss Ephemeris can calculate

that,

> too.)

>

> Penultimately, the ayanamsa. I haven't had much luck using any

linear

> interpolation or extrapolation (Mani is already getting red in the face, I

> know) to estimate it. My linearly-interpolated results rarely coincide

what

> what GJ computes, and I'm not doubting GJ's accuracy here--either my

> reference data is ratty or I'm using too simplistic of an interpolation

> model; one day, I'll try a Lagrange spline. So, instead, I use the Swiss

> Ephemeris to compute the coordinates of Spica (the first point of Libra)

and

> add 180 degs (plus like an arc minute or something) to it. This gives the

> Lahiri first point of Aries. Emotionally, that approach is more

satisfying,

> anyway. If somebody knows what I'm doing wrong, please tell.

>

> And finally, be suspicious about algorithms you find on the web. I

> found one article, supposedly written by an ASTRONOMER no less, who

claimed

> you could determine the lunar naksatra by computing the modulus of [ the

> date in question - some reference date, in days ] by 27.3 (a very rough

> approximation of the lunar sidereal month). The moon's motion just isn't

> that simple that it can be modeled accurately with only three sig figs.

So,

> if it's simple, suspect it: The only thing simple is getting it wrong.

> Have fun, though ... it's a worthy occupation of time.

>

> jpd

>

>

>

> gjlist-

>

>

>

> Your use of is subject to

>

>

>

[Guest guest]

Hey Scott--

 

>Thank you so much for the information. As a rule of thumb I have learned

>not to expect much in this life spoon fed to me, but I did think the

>information would be more readily available on the web than it appears to

be

>(perhaps I'm just not structuring my search strings correctly).

 

Oh, it's there all right, just much of it is WRONG. Seriously, check

out the Schlyter tutorial; I found the URL:

http://hotel04.ausys.se/pausch/comp/tutorial.html

 

BTW, are you the same Scott Fallin of kha0tic Linux security weenydom?

 

 

jpd

[Guest guest]

Hey Phil,

 

> Oh, it's there all right, just much of it is WRONG. Seriously, check

> out the Schlyter tutorial; I found the URL:

> http://hotel04.ausys.se/pausch/comp/tutorial.html

 

Thanks!

 

> BTW, are you the same Scott Fallin of kha0tic Linux security weenydom?

 

Oh no. My past failure has come back to haunt me. I am indeed one and the

same. Bit off more than I could chew there, I'll be the first to admit. I

still hope to make a contribution to Linux security, but on a much smaller

scale.

 

Small world

 

-s

[Guest guest]

Hey Scott--

 

>Oh no. My past failure has come back to haunt me. I am indeed one and the

>same. Bit off more than I could chew there, I'll be the first to admit. I

>still hope to make a contribution to Linux security, but on a much smaller

>scale.

 

You are being modest. Really, that your name is so recognizeable quite

expresses that you've already made a notable contribution to the annals of

our profession. Never make small plans, right? And how many years did it

take for someone to finally harden BSD to the point it was suitable for use

as a firewall? Pretty cool idea, actually; it just hasn't had time to

catch on, I think.

 

>Small world

 

That, indeed, it is. ;-)

 

jpd

 

P.S. Sorry for the off-topic post, but I didn't want anyone getting the

wrong idea about the worthiness of Scott's product.

[Guest guest]

Scott A. Fallin wrote:

>

> I am able to calculate a Western Tropical based chart 'by hand', but would

> like to be able to do the same with a Vedic/sidereal chart. I've yet to

> find such a reference on the net, although surely there is one.

 

 

Phil Draughon wrote:

> Hey Scott--

>

> Penultimately, the ayanamsa. I haven't had much luck using any linear

> interpolation or extrapolation (Mani is already getting red in the face, I

> know) to estimate it.)

 

 

 

Hi!

 

Can´t get red, being very brown, but hope my letters are read!

 

One has to be spoon-fed at some stage or other, unless one calculates the

planetary positions the astronomical way, right from scratch. But even the best

prgrammes use a standard ephemeris, are spoon-fed to that extent.

 

If you can cast a western chart, it is actually the best way to get a vedic or

sidereal chart: subtract the ayanamsa (or add if before say 200 AD!)from all

tropical positions, including house cusps.

 

The reason is that I know of no really reliable sidereal ephemeris. Carl Stahl

used to publish one, but used Fagan´s ayanamsa. Don´t know if it is still

published.

 

The position of the moon is tricky: it should not be simply interpolated between

two midnight positions. One can interpolate between 6 hours, but even that is

not really correct, for the moon rapidly accelerates and decelerates. This

applies to all the bodies, but their daily motion is too small to make any

significant difference within a 3-day period. With the moon, to arrive at a near

precise value the motion during the current 6-hour period has to be corrected

using the previous and successive 6-hour periods.

 

It is not worth the labour these days! If astrology is your hobby, the calcs

ruin your free time - and very likely you make a mistake! If you are a

professional, it takes up valuable time! To do a full vedic chart with all the

vargas, balas etc. you need a day or even two.

 

If you have a reliable programme that calculates fixed star positions, the

ayanamsa can be calculated more accurately than by linear methods. As a

reference star, Spica is not bad, It has a proper motion, but within a 100 years

not much. More accuracy can be had using Shaula. Its tropical position in deg,

min and secs in relation to 0° sagittarius is the Chandrahari ayanamsa. On 1 jan

1900 Shaula was in tropical Sagittarius 23° 11´16". So that was the CH ayanamsa

on that date. Lahiri was 22°27´55", i.e. 43´21" less. So if you find Shaula´s

position for any date and subtract 43´21" from it, you get the Lahiri ayanamsa.

A similar correction can, of course, be made for any other ayanamsa too.

 

But in all basic calcs, use the tropical, astronomical source. That is what all

programmes do too.

 

GJ is the first all-round programme for vedic - parasaras light is perhaps

older, but, I think, less comprehensive. The first vedic programme that was

reliable was Astral-11 for DOS. It was rather limited in facilities, programmed

by a person who did not understand astrology at all, emigrated from India to

Silicon Valley and is perhaps still there! PCJyothish was even more limited.

 

regards

Mani

[Guest guest]

>If you have a reliable programme that calculates fixed star positions, the

>ayanamsa can be calculated more accurately than by linear methods. As a

>reference star, Spica is not bad, It has a proper motion, but within a 100

years

>not much. More accuracy can be had using Shaula. Its tropical position in

deg,

>min and secs in relation to 0° sagittarius is the Chandrahari ayanamsa. On

1 jan

>1900 Shaula was in tropical Sagittarius 23° 11´16". So that was the CH

ayanamsa

>on that date. Lahiri was 22°27´55", i.e. 43´21" less. So if you find

Shaula´s

>position for any date and subtract 43´21" from it, you get the Lahiri

ayanamsa.

>A similar correction can, of course, be made for any other ayanamsa too.

 

Dear Mani,

 

Thanks very much. This is very useful information. Thanks so much for

doing the research for us.

 

jpd

[Guest guest]

Hi Mani,

 

Thank you so much for sharing that information. It is most appreciated.

 

While I have no intention of developing an astrology application, I do like

to know the mechanics behind the subject I'm about to explore as deeply as I

intend to explore Jyotish.

 

Thanks!

 

scott