NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: Rejecting outliers
From: Gary LaPook
Date: 2011 Jan 8, 18:05 -0800
From: Gary LaPook
Date: 2011 Jan 8, 18:05 -0800
I said in my previous post that a navigator might have determined several sigmas based on observing conditions so would have an appropriate sigma to apply to the current series of sights. As to your suggestion to use the sigma determined by the current series of sights, the sigma for my six shots was 0.77' so using my criteria for rejecting a sight of 2.866', which is 3.7 sigmas (of the current series), less than one in a thousand sights should exceed that error so my method might have been overly protective of the data. Of course there could be situations where it would be the other way around so then it might make sense to use the larger of the two sigmas for determining the cutoff value. gl --- On Sat, 1/8/11, George Huxtable <george@hux.me.uk> wrote:
|