Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/3/83 based; site hou2g.UUCP
Path: utzoo!watmath!clyde!burl!mgnetp!ihnp4!houxm!hou2g!stekas
From: stekas@hou2g.UUCP (J.STEKAS)
Newsgroups: net.puzzle
Subject: Re: wind direction measurement
Message-ID: <289@hou2g.UUCP>
Date: Wed, 18-Jul-84 11:57:51 EDT
Article-I.D.: hou2g.289
Posted: Wed Jul 18 11:57:51 1984
Date-Received: Thu, 19-Jul-84 03:21:43 EDT
References: <2596@decwrl.UUCP>
Organization: AT&T Bell Labs, Holmdel NJ
Lines: 11

The correct way to do it is to define a second coordinate system which
is rotated 180 degrees with respect to the initial one.  Now the sample
can be divided into two sub-samples such that each sub-sample contains
measurements between 90-270 degrees in either one coordinate system or
the other. The average direction of each sub-sample can then be calculated
without any problems with the singularity at 0 degrees.  Once that's done
the global average may be calculated by weighting the average of each
sub-sample by the # of measurments in it and adding.  But be sure that
you pick the right (90-270 degree) coordinate system!

Jim