Filtering aggregated values

Introduction

There are 2 ways to apply a widget filter :

on the raw values : each row in the raw dataset is evaluated against the filter condition
on the aggregated values : the raw dataset is first aggregated by the Widget, and only then the filter condition is applied

Example

The Data Source

Imagine a simple dataset :

TagName	Building	Product	Price
FT101.Value	Building A	Product X	10
FT102.Value	Building A	Product Y	15
FT201.Value	Building B	Product X	20
FT202.Value	Building B	Product Y	25
FT301.Value	Building C	Product X	30
FT302.Value	Building C	Product Y	35

The Widget

You are displaying this data in a Pie Chart with Sum(Price) grouped by “Product”, you get :

Product	Sum(Price)
Product X	60
Product Y	75

Filtering raw values

You apply a filter on the raw value to see only data where :
Price > 20

The raw dataset will be filtered like this :

TagName	Building	Product	Price
FT202.Value	Building B	Product Y	25
FT301.Value	Building C	Product X	30
FT302.Value	Building C	Product Y	35

And the Widget will display the following :

Product	Sum(Price)
Product X	30
Product Y	60

(because 60 = 25 + 35)

You apply a filter on the aggregated value to see only data where :
Sum(Price) > 70

It is not the raw dataset that will be filtered.
The Widget will display the following :

Product	Sum(Price)
Product Y	75

(Product X was filtered because its Sum(Price) was 60 <= 70)

How to filter by aggregated values

For each filter condition, you can check Use Field Aggregate :

Then select the desired aggregation function :

Note on non-numerical fields

On fields where the data type is not number (string, date, boolean), the aggregation functions will return a number, so the condition becomes a number condition.
For those fields, the list of available aggregation functions is also different :