Skip to main content

This content has been archived and is no longer being updated.

Links may not function; however, this content may be relevant to outdated versions of the product.

Database tables for monitoring models

Suggest edit Updated on September 10, 2021

Database tables contain monitoring information for adaptive and predictive models. This data is used to populate the charts in the Monitor tab of every model rule in Dev Studio and additional reports in the Actions > Reports section of Prediction Studio.

The tables are transparently designed to contain monitoring data that you can easily use to create your custom reports in Pega Platform™ or any other reporting tool. Before you customize model reports to serve your business needs, learn what properties you can find in these tables.

For more information about how the Adaptive Decision Manager service captures and populates that monitoring data, see Configuring the Adaptive Decision Manager service.

Properties in the PR_DATA_DM_ADMMART_MDL_FACT table

Properties with the py prefix are defined on the Data-Decision-ADM-ModelSnapshot class. The out-of-the-box data set which applies to this class is pyModelSnapshots.

pr_data_dm_admmart_mdl_fact table properties
PropertyDescription
pxApplicationThe name of the application.
pyAppliesToClassThe Applies To class of the Adaptive Model rule with which the model is associated.
pyModelIDA unique identifier for each model.
pyConfigurationNameThe name of the Adaptive Model rule with which the model is associated.
pySnapshotTimeThe time when the snapshot was taken.
pyIssueThe default model identifier that defines the model context in your application. For more information, see Model context.
pyGroupThe default model identifier that defines the model context in your application.
pyNameThe default model identifier. When the context of the models is different from the default hierarchy, this property contains the context values in the JSON format.
pyChannelThe default model identifier that defines the model context in your application.
pyDirectionThe default model identifier that defines the model context in your application.
pyTreatmentThe default model identifier that defines the model context in your application.
pyPerformanceThe performance of each ADM model.
pySuccessRateNot used in reports.
pyResponseCountThe number of all responses for each model.
pxObjClassNot used in reports.
pzInsKeyNot used in reports.
pxInsNameNot used in reports.
pxSaveDateTimeNot used in reports.
pxCommitDateTimeNot used in reports.
pyExtensionNot used in reports.
pyActivePredictorsThe number of active predictors for each model.
pyTotalPredictorsThe number of all predictors for each model.
pyNegativesThe number of negative responses for each model.
pyPositivesThe number of positive responses for each model.
pyRelativeNegativesNot used in reports.
pyRelativePositivesNot used in reports.
pyRelativeResponseCountNot used in reports.
pyMemoryNot used in reports.
pyPerformanceThresholdNot used in reports.
pyCorrelationThresholdNot used in reports.
pyPerformanceErrorNot used in reports.
pyModelDataThe compressed and encoded scoring model.
pyModelVersion

A unique identifier for the model version. When a model is updated, it receives a new version identifier. Every decision also contains this model version, so if Full Auditability is enabled, you can use the model version to find the exact model used.

pyFactoryUpdatetimeThe time when the model factory was last updated.

Properties in the PR_DATA_DM_ADMMART_PRED table

Properties with the py prefix are defined on the Data-Decision-ADM-PredictorBinningSnapshot class. The out-of-the-box data set which applies to this class is pyADMPredictorSnapshots.

pr_data_dm_admmart_pred_fact table properties
PropertyDescription
pxCommitDateTimeNot used in reports.
pxSaveDateTimeNot used in reports.
pyModelIDA unique identifier for each model.
pxObjClassNot used in reports.
pzInsKeyNot used in reports.
pxInsNameNot used in reports.
pyPredictorNameThe name of the predictor.
pyContentsThe overall range of the predictor.
pyPerformanceThe performance of the predictor.
pyPositivesThe number of positive responses for the predictor per model.
pyNegativesThe number of negative responses for the predictor per model
pyTypeThe type of predictornumeric or symbolic.
pyTotalBinsThe number of bins for the predictor per model.
pyResponseCountThe number of responses for the predictor per model.
pyRelativePositivesNot used in reports.
pyRelativeNegativesNot used in reports.
pyRelativeResponseCountNot used in reports.
pyBinNegativesThe number of negative responses for the predictor per model in the bin.
pyBinPositivesThe number of positive responses for the predictor per model in the bin.
pyBinType

The type of bin:

  • Equibehavior a bin type for the normal bins. Bin ranges have the same values within one bin.
  • Missing a bin type for empty values.
  • Residual a bin type for symbols that did not fit in any of the normal bins because, for example, they occur relatively infrequently. This bin type applies to symbolic fields.
pyBinNegativesPercentageThe percentage of negative responses in the bin out of all responses for the predictor.
pyBinPositivesPercentageThe percentage of positive responses in the bin out of all responses for the predictor.
pyBinSymbolThe actual range of the bin.
pyBinLowerBoundThe lower bound of the bin. This property applies to numeric predictors.
pyBinUpperBoundThe upper bound of the bin. This property applies to numeric predictors.
pyRelativeBinPositivesThe difference between the number of positive responses for the predictor in the bin that is measured per model and the number of positive responses for the predictor in the bin from the last snapshot.
pyRelativeBinNegativesThe difference between the number of negative responses for the predictor in the bin that is measured per model and the number of negative responses for the predictor in the bin from the last snapshot.
pyBinResponseCountThe total number of responses for the predictor in the bin that is measured per model.
pyRelativeBinResponseCountThe difference between the number of all responses for the predictor in the bin that is measured per model and the number of all responses for the predictor in the bin from the last snapshot.
pyBinResponseCountPercentageThe percentage of responses falling under a particular bin for a predictor that is measured per model.
pySnapShotTimeThe time when the snapshot was taken.
pyBinIndexAn index assigned to a predictor bin entry. For example, when a model contains the AGE predictor that has 10 bins, then we have 10 entries in the table and the pyBinIndex for each entry is: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10.
pyLiftBehavior in the predictor bin divided by the overall behavior.
pyZRatioThe number of standard deviations in the behavior of the predictor bin differs from the overall behavior.
pyEntryTypeAn indication whether it is an active or inactive predictor. Active predictors are the ones that are used by the model.
pyExtensionNot used in reports.
pyGroupIndex

Not used in reports.

This property cannot be compared between different model versions (updates) because it is just an index.

pyCorrelationPredictorNot used in reports.

Properties in the PR_DATA_DM_BINARY_DISTRIBUTION table (binary outcome predictive models)

For binary outcome models, the count of the positives and negatives is stored in a granular set of bins which are used to calculate the AUC and the ROC curves. The score distribution and the observed responses overlay occur during training.

pr_data_dm_binary_distribution table properties
PropertyDescription
pxApplicationThe name of the application.
pyRulesetNameThe name of the ruleset to which the Predictive Model rule belongs.
pyRulesetVersionThe version of the ruleset to which the Predictive Model rule belongs.
pyModelIDA unique identifier for each model.
pyFactoryKeyA key that corresponds to the version information in the ADM factory data set (internal audit usage).
pySnapshotTimeThe time when the snapshot was taken.
pySnapshotIDAn identifier for the snapshot.
pyBinIDA unique identifier for each bin that is created.
pyBinLabelA label for each bin with ranges.
pyBinUpperThe upper boundary of the bin.
pyBinLowerThe lower boundary of the bin.
pyPredictorNameThe name of the predictor for which the statistic is stored.
pyCountThe total number of responses for the bin.
pyPositiveCountThe number of positive responses.
pyNegativeCountThe number of negative responses.
pyDataUsageThe stage of the model building process at which the data was derivedtraining, predicted, or observed.
pyMetricTypeThe metric type for the binbinary distribution or side-by-side distribution.
pyBehaviorThe behavior of the bin, usually the percentage of positive responses.
pyPercentageThe percentage of responses in the bin.
pxObjClassNot used in reports.
pzInsKeyNot used in reports.
pxInsNameNot used in reports.
pxSaveDateTimeNot used in reports.
pxCreateDateTimeNot used in reports.
pxUpdateDateTimeNot used in reports.
pxCommitDateTimeNot used in reports.

Properties in the PR_DATA_DM_CONTINGENCYTABLE table (categorical outcome predictive models)

For categorical outcome models, the confusion matrix of responses is the main statistic. Each cell in the confusion matrix is stored as a record in the database table.

The confusion matrix is used to calculate the performance values, such as the F-statistic for the model or the accuracy for the classes.

pr_data_dm_contingencytable properties
PropertyDescription
pxApplicationThe name of the application.
pyRulesetNameThe name of the ruleset to which the Predictive Model rule belongs.
pyRulesetVersionThe version of the ruleset to which the Predictive Model rule belongs.
pyModelIDA unique identifier for each model.
pyFactoryKeyA key that corresponds to the version information in the ADM factory data set (internal audit usage).
pySnapshotTimeThe time when the snapshot was taken.
pySnapshotIDAn identifier for the snapshot.
pyIdentifierAn identifier for the category combination.
pyBinLabelA label for each bin with ranges.
pyPredictedCategoryThe category at the make-decision time.
pyActualCategoryA category that was retrieved from the response.
pyDataUsageThe stage of the model building process at which the data was derivedtraining, predicted, or observed.
pyDistributionTypeThe type of distribution that is storedresidual values or actual values.
pyCountThe number of values that are observed in each bin.
pxObjClassNot used in reports.
pzInsKeyNot used in reports.
pxInsNameNot used in reports.
pxSaveDateTimeNot used in reports.
pxCreateDateTimeNot used in reports.
pxUpdateDateTimeNot used in reports.
pxCommitDateTimeNot used in reports.

Properties in the PR_DATA_DM_HISTOGRAM table (continuous outcome models)

For continuous outcome models, the difference between the predicted outcome and the actual outcome is used to measure the performance. The distribution of these residual values is stored in bins of equal interval size. The Information that is gathered in the bins is used to calculate the root-mean-square error (RMSE) and mean absolute error (MAE) performance statistics for the model.

pr_data_dm_histogram table properties
PropertyDescription
pxApplicationThe name of the application.
pyRulesetNameThe name of the ruleset to which the Predictive Model rule belongs.
pyRulesetVersionThe version of the ruleset to which the Predictive Model rule belongs.
pyModelIDA unique identifier for each model.
pyFactoryKeyA key that corresponds to the version information in the ADM factory data set (internal audit usage).
pySnapshotTimeThe time when the snapshot was taken.
pySnapshotIDAn identifier for the snapshot.
pyBinIDA unique identifier for each bin that is created.
pyBinLabelA label for each bin with ranges.
pyBinUpperThe upper boundary of the bin.
pyBinLowerThe lower boundary of the bin.
pyDataUsageThe stage of the model building process at which the data was derivedtraining, predicted, or observed.
pyDistributionTypeThe type of distribution that is storedresidual values or actual values.
pyCountThe total number of responses for the bin.
pyMinimumThe minimum value in the bin.
pyMaximumThe maximum value in the bin.
pyAverageThe average value in the bin. The minimum, maximum, and average values help describe the distribution of data in the bin.
pxObjClassNot used in reports.
pzInsKeyNot used in reports.
pxInsNameNot used in reports.
pxSaveDateTimeNot used in reports.
pxCreateDateTimeNot used in reports.
pxUpdateDateTimeNot used in reports.
pxCommitDateTimeNot used in reports.

Properties in the PR_DATA_DM_SNAPSHOTS table (snapshot summary)

The monitoring information that is stored in the monitoring data mart contains data that is related to the same point in time; that collection of monitoring data is called a snapshot. This table contains one record per snapshot per model.

This information can be linked to data in other tables that contain more detailed, binned information that is used to calculate the performance statistics.

One-time model data may consist of the following details:

  • The number of responses in the snapshot
  • The performance of the model during the snapshot
  • The user that triggered the snapshot
pr_data_dm_snapshots table properties
PropertyDescription
pxApplicationThe name of the application.
pyRulesetNameThe name of the ruleset to which the Predictive Model rule belongs.
pyRulesetVersionThe version of the ruleset to which the Predictive Model rule belongs.
pyModelIDA unique identifier for each model.
pyFactoryKeyA key that corresponds to the version information in the ADM factory data set (internal audit usage).
pySnapshotTimeThe time when the snapshot was taken.
pySnapshotIDAn identifier for the snapshot.
pySnapshotDayThe day when the snapshot was taken.
pyValueThe value of the (meta) property.
pyLabelThe label of the category.
pySnapshotTypeThe frequency of taking the snapshotdaily, weekly, monthly, and so on.
pyDataUsage
  • The stage of the model building process at which the data was derivedtraining, predicted, or observed.
  • The type of group to which the snapshot belongsControl or Test in case of predictions.
pyIdentifierThe identifier of the category for the snapshot.
pyNameThe name of the (meta) property.
pyDistributionTypeThe type of distribution that is storedresidual values or actual values.
pyCountThe total number of responses for the bin.
pxObjClassNot used in reports.
pzInsKeyNot used in reports.
pxInsNameNot used in reports.
pxSaveDateTimeNot used in reports.
pxCreateDateTimeNot used in reports.
pxUpdateDateTimeNot used in reports.
pxCommitDateTimeNot used in reports.

Properties in the PR_DATA_DM_NOTIFICATION table

This table contains Prediction Studio notifications that inform about sudden drops in predictive performance, models with low performance, or other issues with models.

pr_data_dm_notification properties
PropertyDescription
pzinskeyNot used in reports.
pxobjclassNot used in reports.
pxinsnameNot used in reports.
pxsavedatetimeNot used in reports.
pxcommitdatetimeNot used in reports.
pylabelThe name of the model for which the notification is generated.
pyrulesetnameThe name of the ruleset to which the model rule belongs.
pyrulesetversionThe version of the ruleset to which the model rule belongs.
pynotificationtypeThe objective on which the notification is generated. For example, notifications can be based on Performance/Response count.
pxcreatedatetimeNot used in reports.
pxupdatedatetimeNot used in reports.
pytriggertimeThe time when the notification is generated.
pxcreateoperatorNot used in reports.
pyconfigidAn identifier that indicates the model insName or model ID for adaptive models.
pycontextThe values of model identifiers that define the model context for an Adaptive Model rule.
pydescriptionThe notification message that is visible in the user interface.
pymodelidA unique identifier of the model rule.
pymodelreferencekeyA reference to the models of a single Adaptive Model rule.
pymodeltypeThe type of model for which the notification is generated (predictive or adaptive).
pynotificationtypeidA unique ID for each notification type.
pypriorityThe priority for each notification type.
pystatusThe status that indicates whether the user has read the notification.

 

Did you find this content helpful? YesNo

100% found this useful

Have a question? Get answers now.

Visit the Collaboration Center to ask questions, engage in discussions, share ideas, and help others.

Ready to crush complexity?

Experience the benefits of Pega Community when you log in.

We'd prefer it if you saw us at our best.

Pega.com is not optimized for Internet Explorer. For the optimal experience, please use:

Close Deprecation Notice
Contact us