Missing Data Strategy

In statistics, imputation is the process of replacing missing data with substituted values. Incomplete, incorrect, inaccurate and irrelevant data are replaced, modified, or deleted. This is also known as data cleansing or data cleaning.
Figure 54.   Missing Data Strategy properties
Image

You may access this view from a variety of locations. For example, expand Config > Services > AnalyticService, right-click Alerts, clickViews > AX Property Sheet, scroll down and double-click Missing Data Strategy.

Property Value Description
Use This Value check box Enables and disables missing data interpolation for the current value.

When enabled, the framework applies this strategy to all requests.

When disabled, the framework ignores this strategy for all requests.

Aggregation Strategy drop-down list Configures how the framework handles missing trend data (data in a series) when processing analytic requests and one or more records are missing for an interval. It applies when even a single record for an interval is missing. It does not apply to value requests.
 NOTE: If the analytic trend request specifies Interval = none, the framework ignores the Missing Data Strategy. 

Ignore Series ignores the entire series if any record even one interval is missing.

Ignore Point ignores any missing records and aggregates the values in the existing records.

Interpolation Algorithm additional properties Defines the algorithm used to interpolate values for missing values (missing records).

None does not interpolate values for missing records.

Linear Interpolation replaces a missing record by linearly interpolating the missing value using the prior and next records on either side of the missing record.

K-Nearest Neighbor is for numeric, enum and Boolean records. This strategy replaces a missing value by calculating the majority value recorded for the item’s nearest K number of neighbors.

K Value Numeric field editor (default = 1, min =1, max = 30) Defines the number of records used by the configured interpolation algorithm when a record is missing to calculate the interpolated value.