嵌套组中的SSRS总和

Question

I have a table with Year, Month, Product, and number of attempts and number of successful attempts.

Data is grouped by Year, then Month and then Product.

I need to add %success (successful/attempts) by Product, and then for the grouping by Month and then by Year.

Before I had year and month in one field. So I was able to do this using Sum(success.value,"period")/Sum(attempts.value,"period").

But how is this done if there are nested groups?

I did see something about adding the parameter "recursive", but haven't got it to work yet.

Answer 1

Aggregation happens implicitly within the scope of the current level. Normally =SUM(Fields!Amount.Value) is all that is required and Reporting Services handles the aggregation automatically.

If you have the following grouping, here is how the aggregation works implicitly if you just use =SUM(Fields!Amount.Value) :

Year    Month    Product    Aggregation
2013                        Sums all products for all months for 2013
        January             Sums all products for January 2013
                 Widget     Sums all Widgets for January 2013
                 Gadget     Sums all Gadgets for January 2013
        February            Sums all products for February 2013 
                 Widget     Sums all Widgets for February 2013
                 Gadget     Sums all Gadgets for February 2013
2014                        Sums all products for all months for 2014
        January             Sums all products for January 2014
                 Widget     Sums all Widgets for January 2014
                 Gadget     Sums all Gadgets for January 2014
        February            Sums all products for February 2014 
                 Widget     Sums all Widgets for February 2014
                 Gadget     Sums all Gadgets for February 2014

In your case implicit aggregation should work fine regardless of the amount of nesting - Reporting Services will work it all out for you. So this expression should work at every level of grouping:

=Sum(Fields!Success.Value) / Sum(Fields!Attempts.Value)

The SUM will be performed within the scope of the current level of grouping implicitly. If this is not happening, you probably have your grouping operating incorrectly. Have a look at the group properties and see what that level of grouping is actually grouping on; the outer group should be Year, the next group Month and the final innermost group Product.

This doesn't group all Januarys together, it signifies that when that field changes it should print a group footer for the previous group and a group header for the next group. In 2013 when January changes to February it triggers a group change and similarly when January changes to February in 2014. Different Januarys, within the scope of the higher level grouping of the Year.

You need to ensure that your data is in the order required for the grouping, either by using ORDER BY [Year], [Month], [Product] in your SQL statement or by implementing sorting on your tablix. Otherwise fields won't be in the right order and group changes will be triggered constantly as the fields are in random orders.

You only need to define the scope explicitly if you are aggregating outside the current scope. For example, if I want to add the percentage that each row is of the total of the dataset, then I need to explicitly use the dataset as my scope for aggregation, so I would use this calculation in my detail row:

=SUM(Fields!Amount.Value) / SUM(Fields!Amount.Value, "MyDataset")

Note that the recursive parameter is only used for tables that have hierarchical data - that is, parent/child relationships such as an organisational chart of employees.