Wednesday, December 14, 2011 6:38 PM
So I had a huge excel sheet with about 400 columns of various questions (eg. What kind of industry are you in, etc), with discretized responses (1 - Tech, 2-Health Care etc).
I brought it into SQL, but you can't have columns with long winded sentences, so I replaced the column headings with Q1, Q2 etc.
Now I am trying to data mine this, so I created a Data Source View and spent three days coming up with a friendly name for each of these 400 questions.
Imagine my horror when I created a Mining Structure and a model and it reverted to using the Q1, Q2 column names instead of the Friendly ones that show up in the DSV.
Please, PLEASE tell me there is an easy way to force these friendly names to be used in a Structure and all subsequent models. The alternative would seem to be to spend another three days doing the exact same re-naming at the structure level rather than the DSV.
Friday, December 16, 2011 9:08 AMModerator
I am working on this issue. I will update with you soon.
Tuesday, December 27, 2011 9:45 PM
In the Data Source view, are you using sql tables/views (from SQL)?
You can replace them with "Named Query" and select the column as different name/alias.
For example, you can select Q1 column as [WhatnewaliasnameIlike] with below TSQL syntax:
select ..... Q1 as [WhatnewaliasnameIlike] ..
Once the alias for the column is in the DSV, you can then create structure and models based on it.
NOTE: If you only want to just create alias for particular mining model column, you can follow below
Hope this helps.
Meer Al - MSFT
- Marked As Answer by Jerry NeeModerator Thursday, January 05, 2012 10:09 AM