Ordering of Variables or Fields Within Enterprise Miner Environment

[Zachary]

I spend a lot of time setting up my SAS table/datafile to be organized in a very particular way. I like to stack the dependent variables more toward the front of my file with the other explanatory ones more toward the rear. There are also plenty of variables that I like categorizing into groups for ease of working with them.

I am now at a phase of exploring a full implementation within Enterprise Miner. Is there a way for me to tell the Miner environment that I would like to see the variables in their original SAS order? The only other solution that I can think of at this moment is to either deal with the alphabetical order or come up with variable names that have a prefix setting them in the order that I would like.

Thank you,

Zach Feinstein, Statistical Data Modeler

P (952) 838-4289  C (612) 590-4813  F (952) 838-2010

SFM Mutual Insurance Company

3500 American Blvd. W, Suite 700, Bloomington, MN 55431

www.sfmic.com

[jakarman]

EMiner is for designed for mining When you build you Em-project all of that results in a complicated bunch of datasets to be found at the OS-folders.

You can see that is an other layer of metadata information.

It will use some ordering when it makes some sense in the mining process that is not necessary the same the ones in the headers of your dataset. More likely it will not be the same.