Replace (overwrite) an attribute configuration

This endpoint allows you to update the machine learning details for an existing unified attribute, specified by attributeConfigurations id number, in a specified project.

In the body, you supply:

• similarityFunction (required) with a value of COSINE, JACQUARD, ABSOLUTE_DIFF, or RELATIVE_DIFF. The default is ABSOLUTE_DIFF.
• one of numericFieldResolution or tokenizer (required). A value must be specified for one of these parameters, but not both. Options for the tokenizer are DEFAULT, STEMMING_EN, BIGRAM, TRIGRAM, BI-WORD, or REGEX.
• attributeRole (optional) with a value of SUM_ATTRIBUTE for spend or CLUSTER_NAME_ATTRIBUTE for the cluster name, as applicable to the project.
• enabledForMl (optional) Boolean, defaults to false if left blank.

See The Attribute Configuration JSON Object.

For example, the key:value pairs in this body update the configuration for My Unified Attribute.

{
  "similarityFunction": "COSINE",
  "enabledForMl": true,
  "tokenizer": "DEFAULT",
  "attributeName": "My Unified Attribute"
}

You can also include the id, relativeId, and relativeAttributeId in the request body to identify the unified attribute.

On success, this call returns an attribute configuration object with all of the configuration parameter values for the unified attribute.