DM_getForcDate

This function returns a table that contains forecasts from a time series prediction model. The time column is of type DATE. You can use this method only for models that contain an absolute time mapping. Absolute time mapping means that the original time column is of type DATE.

Syntax

Function syntax


>>-DM_getForcDate--(--regModel--)------------------------------><

Parameters

regModel: A value of type DM_RegressionModel

Return Value

The return value is of type TABLE

Table 1. Result Values for DM_getForcDate
Column	Data type	Purpose
submodel	VARCHAR(128)	name of the sub-model that is used to calculate the prediction. Each PMML document can contain multiple time series models for different columns of a given table.
method	VARCHAR(30)	name of the method that is used to calculate the prediction. Valid values are 'ARIMA', 'ExponentialSmoothing', 'SeasonalTrendDecomposition'.
date	DATE(11)	time values of the forecast of type DATE.
value	REAL	predicted value of the time series at time 'date'.
stddev	REAL	predicted standard deviation or incertitude of the prediction. The error value is calculated from the analysis of previous post prediction errors on the training data. Adding or subtracting the standard deviation from the predicted value results in an interval. Based on the assumption that errors are distributed normally around the predicted value, this interval corresponds to the 68%-confidence interval.