Understand rollup analyses

A rollup analysis calculates statistical values such as sum and average for selected attributes associated with an element. For example, a rollup analysis on a factory element might use temperature attribute values for all pumps in the factory to calculate their average temperature.

You can choose either of two sources for attributes to include in a rollup analysis:

• Attributes of the element

  For example, you want to verify that the level of a tank is constant by checking that all inflows and outflow sum to zero. From the list of the tank element's attributes, select its inflow and outflow attributes for a summation calculation. A rollup using an element's own attributes is also known as an aggregation.

• Attributes of the element's child elements

  For example, you want to calculate average energy consumption for a group of pumps in a refinery. To do that, you create a roll-up analysis on the parent element (the refinery) that averages energy consumption attributes from its child elements (pumps).

Use selection criteria to select attributes for rollup. You can select attributes that match attribute name text you enter or that are associated with a category or element template you specify.

With PI Server 2015 R2 and later versions, you can select child attributes for rollup analyses. For more information, see Create a rollup analysis.

You must specify one or more functions to calculate statistics on the selected attributes in the rollup (see Learn about rollup analysis functions). Be sure each output is mapped to an output attribute.

Be aware that:

• You should select attributes carefully to obtain sensible results. Avoid selecting a set with mixed units of measure (UOM; such as temperature and mass) or value types (such as numeric values and time stamps). As long as the inputs and outputs have different units within the same UOM class, the system makes reasonable effort to convert the input units to that of the output for calculation.

• You can add or delete elements or attributes in your hierarchy without the need to update roll-up analyses. Because a rollup identifies input attributes each time it is executed, it automatically includes any new attributes that meet its selection criteria.

• Some selected attributes (such as those with string values) will not participate in rollup calculations, which operate only on numeric or DateTime values.