One of the most common errors faced when configuring Hyperion’s Rolling Forecast functionality in 18.104.22.168 is: “Invalid input parameters were supplied to generate Rolling Forecast.”
This error message can be be triggered due to a number of situations, but before we dig in, let’s review the prerequisites for building automatic rolling forecasts. These requirements are spelled out in full starting on page 149 of Oracle’s official documentation here.
The first step is to design a data form with year and period in the same grid axis of the form. This means both period and year must be in the rows of the form, or both must be in the columns of the form.
Then, using the right click menu on the column, bring up the Rolling Forecast Setup window
Now, begin configuring the rolling forecast to organizational specifications. Let’s look at the example below:
By default, the rolling forecast setup takes the existing column’s year and period, if a single member is selected. For this case, the month of May has been input, as well as a 12 month forecast.
However, when we press generate, we see this will not successfully process the parameters as set, instead displaying the error, “Invalid input parameters were supplied to generate Rolling Forecast.”
The issue in this case is the use of a substitution variable in the Start Year parameter. The Rolling Forecast generation process essentially configures a set of substitution variables, one year and one period each per the number of forecasting periods selected. Inserting a member of the year dimension as a member for start year allows the rolling forecast generation process to proceed.
However, for this application, we’re still not quite there. Upon attempting to generate 12 periods of forecast variables, the system returns a new error, “Insufficient number of years to generate substitution variables for these periods.”
Fortunately this error is very descriptive, and tells us we need to add another year to our application to proceed.
(Note: Remember to refresh your application after adding the new year dimension!)
Finally, with our parameters correctly specified, and all periods and years setup for variable generation, the rolling forecast generation process succeeds, generating a pair of substitution variables for each forecast period specified, as well as adding each combination of period and year into the form being edited.
Have you run into any other errors while developing rolling forecasts? Let us know in the comments!