Missing Value Imputation | Teradata Vantage - 17.10 - Missing Value Imputation - Advanced SQL Engine - Teradata Database

Teradata Vantageā„¢ - Time Series Tables and Operations

Product
Advanced SQL Engine
Teradata Database
Release Number
17.10
Release Date
July 2021
Content Type
Programming Reference
Publication ID
B035-1208-171K
Language
English (United States)

If there are missing values in the data several modes are supported for the imputation of missing values both during the intermediate computation of an aggregate and in the result set for all timebuckets. Notice in the table created in Table and Data Definition for Time Series Aggregates Examples there are some null values for the TEMPERATURE column and not all the timebuckets are represented.

With missing values, consider the following:
  • How the missing values within a timebucket affect the result for that group; for example, should the missing values be ignored, treated as 0, or handled another way.
  • If you want a result for each timebucket, you must specify the manner in which this result should be computed (for example, should the result be ignored, returned with a constant value, and so on).

There are different approaches for each issue presented above which affect the overall result obtained. The following describes the supported modes of operation for all time series aggregates and how to specify the desired behavior in SQL.

Note that only the AVERAGE function is used as an example, but the modes described apply to all aggregates.