Calculate normalized euclidean distance

[Liang Wang Jack]

Hi all,

The Euclidean distance matrix could be obtained by using

use http://www.stata-press.com/data/r10/iris.dta,clear
matrix diss D = seplen sepwid petlen petwid, L2

Whereas, how can I calculate the normalized Euclidean distance according to the formula below?

This formula is coming from the paper http://www.public.asu.edu/~huanliu/p...pakdd00clu.pdf page 4.
The tricky thing is the difference between two observations at a variable k is divided by the maxk - mink value of that variable k before calculating the distance.

Thank you so much for helping!

Best regards,
Jack

[Liang Wang Jack]

Is the normalized Euclidean distance is the Mahalanobis distance? If it is, how could I get the Mahalanobis distance by Stata? Thanks.

[Stephen Jenkins]

You can find out whether you can calculate Mahalanobis distances in Stata by doing a search! Try search mahalanobis and look at the components of the mahapick package , for instance. (Also it could be useful for future reference to consult the FAQ, e.g. https://www.statalist.org/forums/help#before.)