hey there davos,

Thanks for the help!

I added a primary key to my main fact table (called it TRUE_ID) - dsv - but I still can't set up a "measure" which segregates by a dimension. That is to say, I set up my clustering mining model like this (to try and cluster ID_NUM based on TECK_VALUE changing over TECK_ID), but when I process and have a look at the output the clusters are defined only in terms of TECK_VALUE, not in terms of TECK_ID - here.

I think the solution may be for me to set up a new measure rather than just one solely based on TECK_VALUE, be good if I could set up measures based on how TECK_VALUE changes with each TECK_ID.

At the moment from a pivot perspective I have =>

m:TECK_VALUE

ID_NUM TECK_ID=1, TECK_ID=2, TECK_ID=3

1

2

3

Which I can't seem to mine properly because the only measure is TECK_VALUE, and I can't get it to include TECK_ID.

I think what I need is -

ID_NUM m:TECK_VALUE & TECK_ID=1, m:TECK_VALUE & TECK_ID = 2 etc etc

1

2

3

Perhaps?

Still confounded!