Hi, I have explained everything in the below screenprint. Please share your views on how to handle this case..
Note: The module is data module where all the data is imported through csv file.
What you want is a subtotal and Anaplan does not support it so far.
There is a way to replicate this:
You would need separate module which will calculate summary based on user filters. User filters have to be created individually per each column you want to filter.
I don't believe you need a separate module as @KirillKuznetsov stated, just separate line item(s) with the formula: If filter condition then "metric you want summed" else 0. Then you show this on the dashboard/page.
@rob_marshall are you saying to create same number of line items (in same or new module won't matter much) and then referenced them from the input line items?
Apply Summary as SUM on those new line items.
Apply filter on the module and formula line item's SUMMARY will change as per applied filter..
Is it what you are recommending?
The source data will have less dimensions so it's better to separate the two sets of line items by two modules. As per best practices. But i agree that everything can be done in one.
I guess it depends on the filters you are using. If you are using the quick filters within the UX, then I believe you are stuck. If you are using "user-defined" line items (Joey can choose his own filter while Mark can pick his own filter), then you will need to dimensionalize the "input" line item by Users (or a custom users list) as well as the results (meaning the line item logic I spoke about above). So, a quick example that is much less complicated than yours....
Trans data module:
Selection Module dimensionalized by Users or a Custom Users list (*** This is where the module can get really big if you have a lot of users defined in the model. So a Custom Users list might be a good idea if not everyone needs to use this functionality). Make sure the summary on this line item is ANY.
Filtered Module (this is what @KirillKuznetsov was referencing) which is dimensionalized by my Products and my Custom Users List (or Users).
Hope this helps,
Thanks @rob_marshall for pouring in the solution but there is a caveat with the proposed solution i.e. Import module is dimensioned on a fairly large list (more than 500k) and an introduction of another (calc) module with 35+ line items will increase the model size exponentially which we don't wish to get into.
That is fine and understandable, but can you expond a bit more on what the requirement is? For example, how will the users want to search the data? Will it be where a member is greater than a certain number or will it be on master data as I did in the above?
@rob_marshall sorry for turning up late!
Search should be based on the Master Data, means user will select the value from list-type line item and that will be looked into SYS module and basis upon the list-items from SYSTEM module the CALC / REPORT modules will be filtered.
After applying this logic, the issue is that CALC / REPORT modules Summary line is not changing Summary value unless CALC / REPORT modules has that Boolean check in line items. For reference:
CALC module output: based upon filter only 2 lines are the result but Summary of each column is of all lines.
This (Summary calculation based on filtered data) sorts only if I apply the Boolean lineitem as a precheck in each line-item as shown below:
But it is an issue, because if I want to refer this this CALC lineitem to in REPORT module then this source line-item data comes out only basis upon the filter (and not overall).
Hope I am clear with my little long msg.
Kindly note, SYS module and CALC / REPORT modules share same granularity but SYS is not time based.