A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
From: Tony Oz
Date: 2018 Oct 10, 14:46 -0700
It transpires that some calculators (and TI-83+ is one of them) are perfectly capable of doing regressions off-line.
On TI-83+ it is quite trivial: just fill two lists, one with sights' times the other with their HCs, and call the QuadReg L1,L2Y1 built-in operation. You will be presented with a string of coefficients for an ax2+bx+c type expression. If you press the GRAPH button - the quality of the approximation can be visually evaluated.
I guess this is no worse than trying to "average" the sights by hand-drawing a line on a paper. Ofcourse - this is a HUGE departure from a traditional and non-electronic way of doing CN, but...
Right now I compare the HCs obtained by the Law of Cosines formula with those from QuadReg. So far the difference can be explained by data truncation or improper rounding (when entering it to the lists). Numerically I see the difference of ±1 arcsecond.
I wonder if this regression stuff can be used for accounting of the observer's movement?