Spurious constants

This forum contains ProfileStage posts and now focuses at newer versions Infosphere Information Analyzer.

Moderators: chulett, rschirm

Post Reply
PhilHibbs
Premium Member
Premium Member
Posts: 1044
Joined: Wed Sep 29, 2004 3:30 am
Location: Nottingham, UK
Contact:

Spurious constants

Post by PhilHibbs »

ProfileStage seems to be a little over-zealous at identifying columns as constant. For instance, I have a table with nearly 600,000 rows, and all but 20 of these rows have zero in a particular field. ProfileStage says this is 100% constant. Is this supposed to happen? What is the threshhold for constancy?
Phil Hibbs | Capgemini
Technical Consultant
chulett
Charter Member
Charter Member
Posts: 43085
Joined: Tue Nov 12, 2002 4:34 pm
Location: Denver, CO

Post by chulett »

Sorry, somewhat spurious comment here...

What would you rather it say? 99.9967% constant? (I left off a few 6s, btw) I'm guessing it decided to round it up for you. :wink:

Or are you wondering if there is a way to not identify zero as a constant value? To treat it more like a null or unknown value, perhaps?
-craig

"You can never have too many knives" -- Logan Nine Fingers
PhilHibbs
Premium Member
Premium Member
Posts: 1044
Joined: Wed Sep 29, 2004 3:30 am
Location: Nottingham, UK
Contact:

Post by PhilHibbs »

chulett wrote:Sorry, somewhat spurious comment here...

What would you rather it say? 99.9967% constant? (I left off a few 6s, btw) I'm guessing it decided to round it up for you. :wink:

Or are you wondering if there is a way to not identify zero as a constant value? To treat it more like a null or unknown value, perhaps?
The difference between 99.9967% and 100% is much bigger than the difference between, say, 50% and 75%.

My colleague has resolved the problem by looking at the cardinality instead, a cardinality of 1 indicates a true constant field.
Phil Hibbs | Capgemini
Technical Consultant
Post Reply