Let's try again

Question

My previous posting "How to load member for "sales same date last year"?" wasn't clear enough in stating my objective so I'll try again.My outline contains a dimension with each day of the month for this year and prior years. (It isn't and can't be tagged as a 'Time' dimension.)The 'Measures' dimension contains a 'Sales' member. The 'Scenario' dimension contains the members 'Actual', 'Budget' and 'Variance'.Using the Excel add-in, I can easily retrieve 'Sales' for 4/1/2002 in row 2 and 'Sales' for 4/1/2001 in row 3. The add-in does not allow retrieval of these values on the same row. To compare sales for this year to sales last year requires too much manipulation with multiple spreadsheets. And, Analyzer can't do it either.I want to create a new member under 'Measures' titled 'PriorYearSales'. A calc script, if possible, would load this member for 4/1/2002 with the 'Sales' amount for 4/1/2001. Thus the Excel add-in would simply reference 4/1/2002 for 'Sales' and 'PriorYearSales' on the same row. I could even create another new member to calculate the variance between 'Sales' and PriorYearSales'.I'm open to any suggestions.Thanks.Phil

Answer 1

A couple of things:1 - I once had the year, month and date in the dame dimension. In one of my cubes, when I separated the year from the month and date, it reduced the calc time from 3 hours to 20 minutes. It also significantly reduced the space the application needed. I strongly suggest this. 2 - You ought to be able to tag the time dimension with time even though your dates really are not dates. I have ACCTGBBADJ and ACCTEBADJ as dates even though they are not dates. Essbase does not care that a date really is not a date because in my experience the YTD calcs is calcing between a range of members based on the selection.3 – If you separated the year and time – you could split off the last four digits into the year category and easily solve your issue with only one load.Throwing all of this away and trying to solve your issues as specified, I would do this:1 - Load your data twice. The first load would be your 2002 data loaded under current year results.2 - The 2nd load would be the 2001 data load under the PriorYearSales.This can be completed in a couple ways depending on how your rule file is set up. Likewise, in the column with the date, select the properties from the field menu and replace 2002 with 2001 for all occurrences.You can try to set up calc scripts but this is complicated and I would not be the person to help. I hope this helps, it is a pretty easy solution to implement.