Interpolator Input - Teradata Vantage

Machine Learning Engine Analytic Function Reference

Product
Teradata Vantage
Release Number
8.00
1.0
Published
May 2019
Language
English (United States)
Last Update
2019-11-22
dita:mapPath
blj1506016597986.ditamap
dita:ditavalPath
blj1506016597986.ditaval
dita:id
B700-4003
lifecycle
previous
Product Category
Teradata Vantage™
Table Description
input_table Contains the time series whose missing values are to be calculated.
time_table [Optional] If you omit time_table, you must specify the TimeInterval argument. The choice of TimeInterval or time_table affects interpolation, but not aggregation. For details, see the descriptions of the InterpolationType and AggregationType arguments.
count_row_number [Optional] Use only with InterpolationType ('loess'(weights ({constant | tricube}), degree ({0 |1 |2}), span(m))), where m is between (λ+1)/n and 1.

input_table Schema

Each row contains one time point in the series and one or more values.

Column Data Type Description
time_column INTEGER, BIGINT, SMALLINT, DOUBLE PRECISION, NUMERIC, DATE, TIME, TIME(n), TIMESTAMP, TIMESTAMP(n) Time point in time series.
value_column INTEGER, BIGINT, SMALLINT, DOUBLE PRECISION, NUMERIC, CHARACTER, CHARACTER(n), or VARCHAR
For data types CHARACTER, CHARACTER(n), and VARCHAR, the only supported interpolation type is 'constant'.
[Column appears one or more times.] Time point value.
accumulate_column Any [Column appears once for each specified accumulate_column.] Column to copy to output table.

Typically, one accumulate_column is a row identifier, such as 'id'.

time_table Schema

Column Data Type Description
time_column Data type of time_column in input_table [Column appears one or more times.] Time point in time series whose missing values are to be calculated.

count_row_number Table Schema

This table has information about the shorter time intervals into which the original time series (in input_table) has been split. Each row represents one shorter time interval.

Column Data Type Description
id INTEGER or VARCHAR if id is from SeriesSplitter output table; otherwise, any data type Identification number of shorter time interval.

If shorter time intervals are produced by SeriesSplitter function, this value is from split_id_column in SeriesSplitter output table; otherwise, this value is from id column in input_table.

n INTEGER or BIGINT Number of data points in individual time series.