NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: A slope example
From: Andrés Ruiz
Date: 2010 Dec 4, 10:54 +0100
Andrés Ruiz
Navigational Algorithms
http://sites.google.com/site/navigationalalgorithms/
From: Andrés Ruiz
Date: 2010 Dec 4, 10:54 +0100
Peter,
Is only an statistical treatment of the data: average and linear regression.
I have attached for those who haven´t got a spreadsheet, the result.
for French friends: droite de hauteur
Regards
-- Andrés Ruiz
Navigational Algorithms
http://sites.google.com/site/navigationalalgorithms/
2010/12/3 Peter Fogg <piterr11@gmail.com>
Andres wrote:An excel sheet, (maybe works under OpenOffice), used by my program "Lunar Distance.exe" with the data provided by Antoine.It calculates the average time and altitude and the best fit line.
HOW does your program calculate this line, Andres? Is the line derived from the data points? This is what is known as the statistical technique of linear regression, which can derive a straight line from ANY set of data points, no matter how random. The problem with this technique is that the derived line is necessarily a function of the data points, there is no objective reference to judge the derived line against. So one or more outliers can significantly skew the line, can distort the result.
If we wish to analyse a set of observations, we KNOW ( = can easily calculate) in advance the slope of apparent rise or fall over a few minutes. In this case it is -57' over 5-minutes of time.
THEREFORE the rational approach is to compare the pattern of observations against THAT line, rather than deriving the line from the data points.