Attribute Statistics

Object types - Where they help you delineate resource objects into midPont object types.
You can view these statistics by going to Resources > All resources > resource > Schema handling > Object types, and selecting Resource statistics in the actions menu for a particular object type.

Object type mappings - Here, statistics help you to asses the validity of mappings within an object type.
Statistics are available separately for resources, Resource statistics, and for the midPoint side, MidPoint statistics, allowing you to check both sides of mappings.

Value Counts - This value distribution shows how frequently individual values occur. For example, for account status, you can see how many accounts are enabled/disabled.

Cardinality - By comparing the Total values and Unique values tiles, you can check how many values are unique within a dataset. This is a good indicator that an attribute may be suitable for identification or correlation. You can check the Value Counts tab for specific details.

Data quality - Statistics highlight potential issues such as missing or null values, or irregular and inconsistent formats. See the Empty values tile for their count, or check the Value Counts and Value Patterns tabs for specific details.

Value Patterns - Shows common structural patterns found in the data, such as various prefixes and suffixes. This is typically used to identify specific object types during delineation. For example, value patterns make it easier for you to identify administrator accounts that are often prefixed using "adm-".

Searching - A search field is available to quickly locate attributes by name. This is especially useful for object classes with large attribute sets.

Re-calculation - If your source data has changed since the last time you had midPoint show you attribute statistics, you can click the Re-calculate button in the right lower corner of the statistics detail pop-up to renew the statistics based on the current data. You can see the time of the last calculation to the left from the button.

Validating AI suggestions - By viewing individual attributes data, you can validate that the suggestions are representative of the entire data set. For example, you may explore the underlying data to understand why AI has generated a particular script for a mapping.

Determine strong identifier candidates - By identifying attributes with near-total, or total uniqueness, and stable and consistent formatting, you can determine quality candidates for correlation.

Detecting weak suggestions - If you see skewed distributions, high null ratios, or unexpected value patters, you may need to review and adjust an AI suggestion.