NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: 2nd Order Regression limitations
From: Brad Morris
Date: 2018 Oct 20, 15:30 -0400
From: Brad Morris
Date: 2018 Oct 20, 15:30 -0400
Kermit
Please do continue your thoughtful investigation of the merits of linear regression. That goes for Tony as well.
I would think a least squares fit to be better but computationally more expensive. Hence my early comments on this topic. However, after being soundly corrected on the topic, I would like to read more.
So if you don't mind Kermit, please do continue.
Thank you
Brad
On Wed, Oct 17, 2018, 4:02 PM Frank Reed <NoReply_FrankReed@fer3.com> wrote:
Regression games... Isn't this just re-inventing the wheel? The least-squares fix methodology spelled out for decades in the back of the Nautical Almanac already does this the right way. Running regressions on the sights beforehand is just doing it "the wrong way", isn't it??
Frank Reed
