NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: Computer space required to get 1" accuracy
From: Ken Muldrew
Date: 2010 Apr 08, 10:04 -0600
From: Ken Muldrew
Date: 2010 Apr 08, 10:04 -0600
On 8 Apr 2010 at 1:48, Antoine Couette wrote: > Frank, if you have some additional early comments, please let me know > them, just in case you might prevent me from taking a too long cut > instead of a short one. I would not like to re-invent the wheel here > since I actually have NO personnal experience in using arrays. Unless I am misunderstanding this exercise, the whole point is to re- invent the wheel (but using a table of coordinates to specify the perimeter instead of a function with coordinates as variables; this to convince the skeptics that vast and cheap storage has replaced fast, cheap arithmetic as the simplest way to get computers to produce an ephemeris). For those of us trained in analytical mathematics, it's hard enough to accept that brute-force arithmetic solutions trump analytic methods; that we can now scrap arithmetic for look-up tables is too much to bear! ;-) > Any Help/Contribution from anybody most warmly encouraged !!! JPL's DE405 is based on 32 day interpolations using chebychev polynomials with time as a variable. So all you have to do is count the number of chebychev coefficients for each body, for a particular 32 day interval, to determine how many places you will need in your look-up table to maintain 1 arcsecond accuracy with linear interpolation between data points (details left as an exercise for the reader). Ken Muldrew.