Model Preparation/Parameter Estimation | Teradata Vantage - Model Preparation and Parameter Estimation Functions - Teradata Vantage

Database Unbounded Array Framework Time Series Functions

Deployment
VantageCloud
VantageCore
Edition
Enterprise
IntelliFlex
VMware
Product
Teradata Vantage
Release Number
17.20
Published
June 2022
Language
English (United States)
Last Update
2024-10-04
dita:mapPath
ncd1634149624743.ditamap
dita:ditavalPath
ruu1634160136230.ditaval
dita:id
ncd1634149624743

Model preparation and parameter estimation functions help data scientists develop forecast models.

You can apply these functions to any discrete series or matrix of data collected at equally spaced points in time or space. That is, any series whose associated indexes are discrete or any matrix whose row and column indexes are discrete.

You can build an ARIMA forecasting model only with a time series that is stationary with respect to mean, covariance, and variance. See Creating an ARIMA Model.

Functions:

TD_ACF
Calculates autocorrelation or autocovariance of time series.
TD_ARIMAESTIMATE
Performs parameter estimation for seasonal and nonseasonal auto-regressive models, moving-average models, and combined ARMA models as ARIMA models.
TD_ARIMAVALIDATE
Performs model validation on a series that was previously processed using TD_ARIMAESTIMATE with FIT_PERCENTAGE less than 100.
TD_DIFF
Applies non-seasonal, seasonal or both types of differencing to the input series to transform a non-stationary series into a stationary series.
TD_LINEAR_REGR
Performs a multivariate linear regression using a formula defining relationship between explanatory variable and multiple response variables to fit data to multidimensional surface. This function is typically used to support the multivariate regression with an ARIMA errors modeling strategy.
TD_MULTIVAR_REGR
Performs a multivariate linear regression using a formula defining relationship between explanatory variable and multiple response variables to fit data to multidimensional surface. This function is typically used to support the multivariate regression with an ARIMA errors modeling strategy.
TD_PACF
Measures the degree of correlation between time series sample points when the effects of the time lags have been removed.
TD_POWERTRANSFORM
Applies power transform to time series with nonstationary variance, creating new time series with stationary variance.
TD_SEASONALNORMALIZE
Normalizes the cyclic, seasonal patterns in the dataset to transform a nonstationary series into a stationary series.
TD_SMOOTHMA
Applies smoothing function to time series with nonstationary mean, creating new time series with stationary mean.
TD_UNDIFF
Reconstructs a differenced time series created by the TD_DIFF function.
TD_UNNORMALIZE
Reverses the seasonal normalization done by the TD_SEASONALNORMALIZE function.