You should be able to get pretty close to this using summaries with hidden key fields. How "pretty" do you need the formatting to be? Page breaks might be trickier, but there's a hack using a two pass solution where you export the summary as fixed width text, and can then define a Page Header template in a second model.
There are some Cumulative Sum measure calculations you can use in the summary.
If these don't get you want you want, you might find it helpful to create some calculated fields.
Assuming that each row of data has fields like:
Branch, Year, Month, Sales
A, 2011, 01, $100
A, 2011, 02, $150
then you can create a field called Sales YTD which is just:
which you can then use as a measure.
That is a similar report layout but each individual branch is listed only once with a YTD total accumulated from all the previous months MLY branch totals. Each branch would then have that YTD total on that report just right below the MLY report. Therefore I am accumulating totals by each individual branch as well as a total amount for all branches combined.
Each page lists the MLY branch report totals as well as the YTD branch report totals for any give month.
I will send you a PDF of the January and February reports to clarify this further.
I would then like to know your ideas on how to accumulate each YTD branch total and how to display both the MLY and YTD figures for each month on one page. I currently have 2 summaries. One for MLY and then one for YTD.
I quick thought here.
If you capture MTD/YTD figures the are fixed data that summarised a set of data as it existed at the tome the report was created. The report program would, presumably, have calculated the numbers form the sum of the detail lines.
If you convert the report back to data using Monarch it afford you and your management to investiogate the newly freed data using different filters and gouopings and, liklely, a load of stuff that was simply not readily available in the original report. Monarch's summaries can, usually, recalculate the revised MTD/YTD figures based on the details in the selected subsets and so on. Of course this could just as well apply to the entire data set too.
Having the previous fixed MTD and YTD numbers is likely useful in some form for model/data verification work - but maybe less useful for further analysis efforts? If you report Detail only goes down to MTD per branch (for example) them the YTD should still be available by calculation.
In other words is this a problems you could step around rather than address head-on?