Finding distribution in a list of values

September 18, 2008

I had a friend of mine ask me a question not long ago. He had developed a site for posting ads on selling houses for a certain company. The managers of the company were now trying to analyze the data on people posting ads. One of the things they were asking him to do, was give them a graph on how many people were posting ads on a certain span of prices. They weren’t going to define him the span of money, but rather wanted him to give them the span, according to the amount of people that existed in it. And that’s were he got stuck… because there was really (almost) no end to how much money can be put on a house, and there seemed to be endless values to choose from in between.

Now I remembered from the Analysis Services Tutorial I did, (which is a great place to start learning about SSAS), that you can group attribute members with the DiscretizationBucketCount property along with the DiscretizationMethod property. The DiscretizationMethod helps you determine the method by which SSAS will group your values (for instance, by Equal Areas) and the DiscretizationBucketCount determines the amount of groups that will be set.

So what was my suggestion? Turn the Sales table into a Fact – Dimension sort of table, where the price of the house is an attribute, and the amount of ads was the measure. We can then use DiscretizationMethod and DiscretizationBucketCount on the attribute of price to help us look at it more clearly.

Add comment
facebook linkedin twitter email

Leave a Reply

Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>