3 months purchase value”. You are creating this attribute to capture the total cart checkout value for the “last 3 months”. In the process, you are going to do a “summation” to aggregate the revenue associated with the “Order_Completed” event name. You can further build the condition for the “Order_Completed” event name using the values from its mapped Zeotap Catalogue Field. These values can be associated with a product category like brand = Puma.
Now that you have a fair understanding of the sample scenario, perform the following steps to create a calculated attribute:
Select the appropriate Aggregation type using the drop-down menu. The values can be 
Average, Count, First Timestamp and more.In Select a field, choose the attribute on which the calculation works based on the chosen aggregation type. Note that the fields available for Aggregation depends on the type of Aggregation choosen in the first step.
You can further specify conditions for your Calculated Attributes by clicking Add Condition. Note that you can also include Non-customer entity data in the conditions based on your requirement. 
In the Group by field, select one or more attributes (up to 3) to define how the aggregated values should be grouped. Group by helps you break down one calculated attribute into multiple values.Example: Instead of calculating a single “3 months purchase value”, you can group it by Product Category to get the total purchase value separately for each category per user.Only attributes that are at the same level as the selected aggregation field are shown for Group by.
For example, if the aggregation is performed on a property inside an object, then only the other properties of that same object (and any joined Non-Customer Entity attributes linked to it) will be available for Group by. Attributes from unrelated objects or levels will not be available.