Understanding Functions in MicroStrategy
MicroStrategy Functions Reference
Additional examples of functions in expressions
© 2011 MicroStrategy, Inc.
drop table ZZT1Y01007DMD001
• Pass0 and Pass1 are issued by Intelligence Server to compute the metric
Daily Sales in 2002. The Intelligence Server prepares the temporary table
with Call Center and Day as its key. Then, it retrieves the sales in 2002
using the transformation Last Year.
• Pass3 and Pass4 are issued to compute the metric Daily Sales in 2003.
Attributes Call Center and Day are used as keys to the temporary table.
This pass is similar to Pass0 and Pass1 with the only difference being that
there is no transformation for this metric.
• In Pass5, the Intelligence Server computes the p-value for each call
center, using the HeteroscedasticTTest and the HomoscedasticTTest
functions, and inserts the values back into the table.
• The remaining passes drop the temporary tables.
In the resulting report, Hypothesis Testing shown above, only San Francisco,
Fargo, and Memphis have a p-value of less than 5%. This indicates that the
probability of making an error in concluding that the sales have significantly
decreased is low. This is strong evidence that average daily sales in San
Francisco, Fargo, and Memphis for 2003 are actually different from those in
Confidence Level example
The reports and report components in this example can be found in
the following folder:
MicroStrategy Tutorial\Public Objects\
Reports\Technical Reports\Reports by Feature\
Analytics\Statistics and Forecasting\
Who are my valuable customers? (Definition 1)
The basic goal is to define a cut-off value that represents the minimum
requirement to be classified as a valuable customer. To identify the valuable
customers in your customer base, you must determine the parameters that
help differentiate those customers from the others. In this example, valuable
customers are those whose spending averages above an upper bound of sales
The ORDER_FACT table contains all orders received in 2002 and 2003.
Assume that the sales order amount is normally distributed with a certain
mean and standard deviation. Based on the assumption of normal