I'm starting an online analytical processing (OLAP) project and the data files are in an Excel spreadsheet. I'm designing an application for the entire OLAP server, but need to know what things I should be considering. How do I know which of the aggregations should be computed and which should be left on fly?
I'm always interested primarily in the data underlying any
online analytical processing (OLAP)
project. When you say the source is Excel, I think about how difficult it is to get users to depend on a locally owned and inconsistently changed spreadsheet. You will want to make sure you have repeatability in the underlying data and can express that to the anticipated users. As for aggregations, and derivations for that matter, my rule of thumb has to do with usage and performance. I think about if the users can easily get to what they ultimately need. If it's a straight rollup (summarization), the OLAP tool should handle that pretty easily. However, if it's going to be a frequently used summary, you still may want to do it as part of the OLAP build cycle. Occasionally, I'm convinced
these summarizations are needed, but usually there is value in having the detail available through OLAP as well. Certainly, if you are deriving new fields for the OLAP, bring over the underlying components of the calculation as well.
Do you have a question about your OLAP project? Ask our expert!
Dig deeper on Business intelligence best practices
Are business intelligence certifications worth it? Get certification advice from our business intelligence expert, William McKnight.continue reading
There are loads of business intelligence tools out there on the market today. Our business intelligence expert provides some valuable resources for ...continue reading
Get examples of how data mining is used in vertical industries, such as retail, manufacturing, healthcare, financial and telecommunications.continue reading
Have a question for an expert?
Please add a title for your question
Get answers from a TechTarget expert on whatever's puzzling you.