Compute distance to centroid

Question

Dear All,

How to compute the distance to all cluster centroids?
For example:
1. apply k-means clustering
2. obtain k centroids
3. for each example in the example set:
compute k distances to k centroids

Best regards,

Wessel

frito · Answer

wessel  wrote:
Hey,

It works, but it is rather slow.

Best regards,

Wessel

hi W
I do this kind of post or pre processing outside of RM in tools that are much more fit for that, I personally use Matlab, 
but since this is algorithmically very simple problem you have, it can be done very fast even in excel. (python and perl should do this in a matter of milliseconds by means of very simple 1 loop script even for 10 of 1000s examples)

if you need to do this on the fly you may want to look at Rapid Analytics where u can run your current RM k-means clustering process and 'chain' it with an external script.

I am afraid RM will always be S L O W with loops. Unless u can write your own operator ;)

br
f

wessel · Answer

Hey,

It works, but it is rather slow.

Best regards,

Wessel

frito · Answer

hi w
so is ur best attempt working?