NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: Multi-Moon line exercise in 2 parts
From: Peter Hakel
Date: 2009 Aug 8, 09:29 -0700
From: George Huxtable <george@hux.me.uk>
To: NavList@fer3.com
Sent: Saturday, August 8, 2009 6:36:14 AM
Subject: [NavList 9412] Re: Multi-Moon line exercise in 2 parts
Peter Hakel (PH) wrote, in [9409], about Jeremy's closely-spaced sequences
of Moon altitudes, in [9359]-
"An important detail is that I obtained this result by disregarding the
second measurement (at 8:54:46) because its intercept does not seem to
"belong" among all the others (it is AWAY, while all others are TOWARD). If
I do include this outlier, the results get worse:"
Yes. A simple plot of Jermy's data "Moonlines, Moon away from LAM" shows
that the observation at 8.54.56 is a real misfit, and should be discarded.
It seems most likely that there was a recording or transcribing error, and
that it should have been noted as 8.52.56, but it's not safe to guess at
such things, and the only safe procedure is to delete that line. Peter did
the right thing.
What, by the way, do the letters LAM stand for?
With that line deleted, those two data sets show remarkably little scatter.
A least-squares fit to the non-culminating series arrives at a rms scatter
of little more that 0.1 arc-minute, and of the series around culmination of
about 0.15 arc-min. These show only random errors, not systematic ones such
as those due to index error observation or dip; they show what can be
achieved by a careful observer on a large vessel in, presumably, benign
conditions.
George.
contact George Huxtable, at george@hux.me.uk
or at +44 1865 820222 (from UK, 01865 820222)
or at 1 Sandy Lane, Southmoor, Abingdon, Oxon OX13 5HX, UK.
--~--~---------~--~----~------------~-------~--~----~
NavList message boards: www.fer3.com/arc
Or post by email to: NavList@fer3.com
To , email NavList-@fer3.com
-~----------~----~----~----~------~----~------~--~---
From: Peter Hakel
Date: 2009 Aug 8, 09:29 -0700
I find it interesting that a single outlier among 30 data points can have such a significant effect on the final result. Clearly, a good software program expected to function as a "black box" (i.e. without needing human intervention to filter out bad data points on input) must carry out the statistics very carefully - and, presumably, display a warning message.
Peter Hakel
Peter Hakel
From: George Huxtable <george@hux.me.uk>
To: NavList@fer3.com
Sent: Saturday, August 8, 2009 6:36:14 AM
Subject: [NavList 9412] Re: Multi-Moon line exercise in 2 parts
Peter Hakel (PH) wrote, in [9409], about Jeremy's closely-spaced sequences
of Moon altitudes, in [9359]-
"An important detail is that I obtained this result by disregarding the
second measurement (at 8:54:46) because its intercept does not seem to
"belong" among all the others (it is AWAY, while all others are TOWARD). If
I do include this outlier, the results get worse:"
Yes. A simple plot of Jermy's data "Moonlines, Moon away from LAM" shows
that the observation at 8.54.56 is a real misfit, and should be discarded.
It seems most likely that there was a recording or transcribing error, and
that it should have been noted as 8.52.56, but it's not safe to guess at
such things, and the only safe procedure is to delete that line. Peter did
the right thing.
What, by the way, do the letters LAM stand for?
With that line deleted, those two data sets show remarkably little scatter.
A least-squares fit to the non-culminating series arrives at a rms scatter
of little more that 0.1 arc-minute, and of the series around culmination of
about 0.15 arc-min. These show only random errors, not systematic ones such
as those due to index error observation or dip; they show what can be
achieved by a careful observer on a large vessel in, presumably, benign
conditions.
George.
contact George Huxtable, at george@hux.me.uk
or at +44 1865 820222 (from UK, 01865 820222)
or at 1 Sandy Lane, Southmoor, Abingdon, Oxon OX13 5HX, UK.
--~--~---------~--~----~------------~-------~--~----~
NavList message boards: www.fer3.com/arc
Or post by email to: NavList@fer3.com
To , email NavList-@fer3.com
-~----------~----~----~----~------~----~------~--~---
Get a NavList ID Code
A NavList ID Code guarantees your identity in NavList posts and allows faster posting of messages.