Let's Get Social!
Focal Point    Focal Point Forums    Performance Management Framework (PMF)    Aggregation of Measures stored as percentages
 Go New Search Notify Tools Reply
 Aggregation of Measures stored as percentages
Member
 posted August 17, 2010 01:25 PM
I am loading industry averages into the "Benchmark" field for a measure. These are stored in a file as percentages. For example, the industry average for cost of goods sold is 62% of Sales. I have a file which contains actual COGS as a percent of Sales, and the Benchmark COGS as a percentage of sales. I want these to roll up from month to quarter to year using the Percentage method. My question is this - does every measure which uses the Percentage aggregation method need to have a Numerator and Denominator loaded? That appears to be the way it works. So if the data is already stored as a percentage, do I need to set up a denominator of "1" to get the roll ups to work correctly? This method works, just wondering if I am missing a simpler solution.

WebFOCUS is 7.6.11
Windows
all output (Excel, HTML, PDF)

 Posts: 6 | Registered: July 22, 2010 IP
Master
 posted August 17, 2010 01:44 PM Hide Post
You want to be able to have the percents roll up, and in a way that respects their percent-ness.

Ideally you'd have access to the source data used for calculating the percents, and then you'd be able to define the numerator and the denominator for the percent measure and let PMF calculate it. That's the mathmatically recommended way to do it.

However if you can't get access to the data used to calculate the percent (for *now* we hope - hopefully this won't be a *permanent* state of affairs) you can take a mean average for aggregation by using a fixed amount in the denominator - 1 for example represents a *single count of percents being aggregated*. Though not a mathmatically perfect approach it will work well enough. Essentially you're saying "mean average these percents for me when I aggregate." A mean average of course being "divide the total by the count of instances."

This is the best you can do mathmatically in any case - either with or without PMF. There's no math magic in the world possible to deconstruct percents back to their components if you don't have access to those components.

thanks

This message has been edited. Last edited by: Bob Jude Ferrante,

Bob Jude Ferrante