Exploring Memory Analysis Views > Exploring the Merged Allocation Points View > Understanding the Merged Allocation Points View
The Merged Allocation Points view is a good starting point for investigating where instances originate. Each method that allocates instances of the selected class becomes the root of a tree. The tree represents the merged stack trace for the method. A merged stack trace includes all methods that called that particular allocating method. Beside each allocating method you can see the number of instances allocated by the method in the Count column.
Tip You may see the same method twice in the list. Duplicate allocating methods occur when two lines of code in the method each allocate instances.
When you expand a stack trace, you can see the series of method calls that led to the allocating method being invoked. The arrows beside the method names show the direction of the calls toward the allocating method. Where more than one series leads to the allocating method, each series is shown as a branch in the tree. The value in the Count column for the allocating method is the total of all branches.
If the Merged Allocation Points view shows “Empty dataset: No matches” either the class has no instances of traces, or allocation point data was not collected at the time that the instances of the class were allocated. Stack traces are only collected on instances that were allocated after you started the use case.