topic Re: How to create a distance matrix in JMP PRO 13 in Discussions

How to create a distance matrix in JMP PRO 13

connquest — Thu, 08 Mar 2018 16:58:37 GMT

I have a sall data table and I'm trying to figure out if there is a way to have JMP Pro 13 create a distance matx against some raw data and then for the standardized data to see if normalizing the data changes which two records are farthest from each other in termsf Euclidean distance? Can someone point me in the direction of where I go in the JMP tool to have it generate the distance matrix?

Re: How to create a distance matrix in JMP PRO 13

cwillden — Thu, 08 Mar 2018 17:25:32 GMT

There's a scripting command for that, Distance(). Check out the documentation in the Scripting Index (Help > Scripting Index).

The trick will just be using the same vector or matrix for both arguments. Here's some examples:

//1-D example
x = [1, 2, 3, 4];
D = sqrt(Distance( x, x )); 

//2-D example
x2 = [1 1, 2 2, 3 3, 4 4];
D = sqrt(Distance( x2, x2 ));

If your data is in a table, see the commands Get As Matrix (to get an entire data table as a matrix) and Get Values (e.g. col1 << Get Values) to get a specific column as a vector. You can also find these in the Scripting Index and Spripting Guide (Help > Books).

Re: How to create a distance matrix in JMP PRO 13

SamGardner — Fri, 11 Aug 2023 16:42:22 GMT

This is an older post, but doing some research on some distance based methods, and came across this. In more recent versions of JMP, you can create a distance matrix table using the Hierarchical Clustering platform, available as a command under the red triangle menu. Here is a JSL example that does an example.

Names Default To Here( 1 );
dt = Open( "$SAMPLE_DATA/Birth Death Subset.jmp" );
obj = dt << Hierarchical Cluster(
	Y( :birth, :death ),
	Label( :country ),
	Save Distance Matrix
);

Re: How to create a distance matrix in JMP PRO 13

ron_horne — Fri, 11 Aug 2023 22:43:05 GMT

hi @connquest ,
in the past we had this discussion:
https://community.jmp.com/t5/Discussions/Distance-Calculation-using-Latitude-and-Longitude/m-p/19258#M17562
which is useful in some application.