LISTSERV at the University of Georgia
Menubar Imagemap
Home Browse Manage Request Manuals Register
Previous messageNext messagePrevious in topicNext in topicPrevious by same authorNext by same authorPrevious page (July 2001)Back to main SPSSX-L pageJoin or leave SPSSX-L (or change settings)ReplyPost a new messageSearchProportional fontNon-proportional font
Date:         Tue, 31 Jul 2001 11:11:43 -0400
Reply-To:     Arthur J Kendall <KendallA@GAO.GOV>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Arthur J Kendall <KendallA@GAO.GOV>
Subject:      Re: euclid distance again
Comments: To: langefeld@TM.TNO.NL
Content-Type: text/plain; charset=US-ASCII

your example data shows only 1 variable for distances but each case has a distance from every other case.

What do you mean by "Euclidean distance" when there is only 1 dimension? With a single dichotomous variable the Euclidean and city block distances are the same. the Euclidean distance is the sqrt of the sum of squared distances on each of the dimensions. d(1,2)=sqrt( ((a1-a2)**2) + ((b-b2)**2) + ((c1-c2)**2) ...) the city block distance is the sum of the absolute distances on each of the dimensions. for the following matrix is for group 1. for the first row (across) for 1 vs 1 the distance is 0 for 1 vs 2 the distance is 0 for 1 vs 3 the distance is 0

1 2 3 4 5 1) 0 0 0 1 1 2 ) 0 0 0 1 1 3 ) 0 0 0 1 1 4) 1 1 1 0 0 5) 1 1 1 0 0

lets say you have 30 subjects that would give you a 30 by 30 matrix with the upper & lower triangles equal to each other. I don't know if PROXIMITIES will work with only 1 variable but if it does you could do the whole 30 by 30 matrix and just copy the submatrices you are interested in. (1 to 5)by (1 to 5) for the first group (6 to 9) by (6 to 9 ) for the second group ...

numeric subject (f3) compute subject= (10 * grade) + pupil.

PROXIMITIES sex /MEASURE = Euclid /id =subject.

>>> Anja Langefeld <langefeld@TM.TNO.NL> 07/31/01 10:20AM >>> Hello,

I will try to explain a bit more clearly.

The issue is this: Let's say I have five groups (e.g. grade school years) of unequal size . The members of each group may be divided into sex (male, female) (see the example matrix below). THe question is what is the euclid distance of each pupil from the others in his or her grade year in terms of the variable "sex"? Is it possible to allow SPSS to compute the euclid distance for each pupil or am I resigned to do it by hand?

grade pupil sex euclid 1 1 1 ? 1 2 1 ? 1 3 1 ? 1 4 0 ? 1 5 0 ? 2 1 1 ? 2 2 0 ? 2 3 0 ? 2 4 0 ? etc.

greetings Anja


Back to: Top of message | Previous page | Main SPSSX-L page