Output Selector: Value from Highest

The Value from Highest output selector evaluates a numeric attribute on each record to select a value to use for another attribute.

Use the Value from Highest output selector where records have a numeric attribute which can be used to select the best record to use for various other output attributes.

This output selector may be useful where you need to use complex logic to pick the best record to use as the merged output record for a number of records representing the same entity. Complex logic, applied before matching, can be used before to create an populate a numeric 'selection score' attribute representing an overall view of the completeness, appropriateness and likely relevance of each record. That score can then be used during merging to choose the best record.

In other cases, there may be a simple Number attribute already in the data that you may want to use in selection - for example, an attribute representing the number of products to which an individual has subscribed.

The following table describes the configuration options:

Configuration Description

Configuration	Description
Inputs	For the actual output `Value`, any input attributes from any input data sets may be configured as long as they share the same data type. A Number attribute must be configured as the `Check Value Attribute` for the actual 'Highest' selection logic.
Options	Specify the following options: `Use first non-empty value if tied?`: this option provides a way of selecting a value arbitrarily if more than one record in the group has the same highest value in the check attribute. The first alphabetically sorted value will be selected for String values, the lowest value for Numbers, and the earliest value for Dates. Type: Yes/No. Default value: Yes. `Consider Nulls as 0?`: this option drives whether to consider Null values as 0, or whether to ignore them. If Nulls are ignored, any numeric value will be considered higher than a Null value. Type: Yes/No. Default value: No.

Inputs

For the actual output Value, any input attributes from any input data sets may be configured as long as they share the same data type.

A Number attribute must be configured as the Check Value Attribute for the actual 'Highest' selection logic.

Options

Specify the following options:

Use first non-empty value if tied?: this option provides a way of selecting a value arbitrarily if more than one record in the group has the same highest value in the check attribute.

The first alphabetically sorted value will be selected for String values, the lowest value for Numbers, and the earliest value for Dates.

Type: Yes/No. Default value: Yes.
Consider Nulls as 0?: this option drives whether to consider Null values as 0, or whether to ignore them. If Nulls are ignored, any numeric value will be considered higher than a Null value.

Type: Yes/No. Default value: No.

Example

In this example, the Value from Highest output selector is used to select an Occupation value of the 'best' record based on a SelectionScore attribute that has been added to all input records before matching.

Example configuration

Use first non-empty value if tied? = No

Consider Nulls as 0? = No

Example output

The following table shows example output using the Value from Highest selector:

Table 1-111 Example Output Using Value from Highest Selector

Input A (Occupation, SelectionScore)	Input B (Occupation, SelectionScore)	Output value(Value from Highest)
CEO, 45	Chief Executive Officer, 60	Chief Executive Officer
Unknown, 0	Account Manager, 60	Account Manager
Sales Executive, 60	Partner Manager, 60	Selection error (requires manual resolution)
Nurse, 45	Chief Nurse, 60	Chief Nurse
null, 0	Secretary, 50	Secretary
Associate Analyst, null	Business Analyst, 20	Business Analyst