Audience Criteria

This section explains about configuring the parameters like timeouts, query limit, bucket size for chart, and reporting options for audience criteria.

"audience": {
            "clientInactivityTimeoutInMillis": 600000,
            "maxNumOfActiveQueriesCached": 10,
            "maxNumOfBucketsForSparseNumericalHistograms": 10000,
            "maxNumOfConcurrentQueriesUnderExecution": 10,
            "maxNumOfEvaluationErrorsToReport": 10,
            "maxNumOfUniqueValuesForCategoricalHistograms": 200000,
            "queryExpirationTimeoutInMillis": 3600000,
            "queryFilterSamplingRate": 1.0
        },

clientInactivityTimeoutInMillis: set the maximum idle time allowed before a client is considered inactive. During the event execution, when the browser is idle for the set time period, the system will throw timeout message.

maxNumOfActiveQueriesCached: set the maximum number of active queries that can be cached.

maxNumOfBucketsForSparseNumericalHistograms: set the maximum number of buckets that can be generated in the numerical histogram chart.

maxNumOfConcurrentQueriesUnderExecution: set the maximum number of concurrent queries that can executed at once. Let assume, the set number of concurrent queries be three, system will execute the first three queries, and execute the fourth query when one of the three queries get completed.

maxNumOfEvaluationErrorsToReport: set the maximum number of evaluation errors to be listed in the report following audience criteria evaluation. The report contains error details of the condition and warnings.

maxNumOfUniqueValuesForCategoricalHistograms: set the maximum number of segments or categories that can be generated from unique values in a histogram chart.

queryExpirationTimeoutInMillis: set the query execution timeout threshold in milliseconds.

queryFilterSamplingRate: set a filter sampling rate within the range of 0 to 1 to sample the records from each partitions and execute the audience criteria. For example, if you set the sampling rate as 0.5, the system will sample half the records from each partition and perform the audience criteria evaluation in each partition.