Working with data from Multiple Processes 

Multiple Processes and distribution by process
When multiple process definitions exist, a distribute by process option is always available.
In some cases this is the only option in the distribution menu. 
Distribution: None =  Multiple processes behave as if a single process. Data is 'compiled'. 
Distribution: Process = Distribute by process 

Distribute by process and distribution by other group is mutually exclusive. 

Configurations: Heatmaps
Heatmaps are not possible for multi process reports. 

Configurations: Multi measure and distributions 
Multi measure, multi process  Reports can only be viewed as a table when distributed. (at this time)

Single measure Reports can be viewed as the standard visualisations 

Configurations: Variables  
All variables existing across processes are represented in the list. 
Variables existing across processes have a single option in the list, and when distributed by process, it is possible to see the value per process. 
There is no hint as to which variables are available across processes at this time. 

Non Camunda engine processes
External event processes & Zeeby processes are missing some data required to use all the report builder configuration options. 
Concept: Allow all selections - in some cases there will be no data for the process that is missing it - this may only become clear when the distribution is set to 'Process' 

Current no data implementation: 
In single process reports visualised as table we display the empty table outline along with a no data message.
In reports visualised as chart - simply no data points are displayed. 

For Multi processs reports distributed by process:
Table = Display empty table columns (we only show the no data in single report cases because there is no column/row number to display and don't want the empty space to be misleading)
Bar = No data points 

At this time, do not inform the user as to why data is missing, though process origin does appear when selecting a non Camunda engine process.  

Process instances and Relative counts 
Believe we disabled it for Combined reports when the data was nonsensical. 
We double process instances in cases of multiples of a process, is it useful to display the relative value of an individual data point relative to the total process instances of all the processes?  Discuss