Time Series Forecasting and Analysis Functions
Time series SQL functions are either OLAP-style aggregate functions, or scalar functions that support OLAP-style aggregates. Use the time series and forecasting functions to use autoregressive integrated moving average
(ARIMA) or generalized autoregressive conditional heteroskedasticity (GARCH) modelling, and to apply model construction and evaluation utilities, such as a Box-Cox transformation.
- Aggregate Time Series Forecasting and Analysis Functions
Aggregate time series forecasting and analysis functions are OLAP-style aggregates designed exclusively for financial time series statistical analysis. As aggregates these functions accept many sets of input values and return a single result.
- Scalar Time Series Forecasting and Analysis Functions
Scalar time series and forecasting UDF functions support the TS_GARCH and TS_AUTO_ARIMA aggregate functions. TS_GARCH and TS_AUTO_ARIMA each produce a binary composite, but also accept binary inputs. The TS_INT_ARRAY function provides inputs for TS_AUTO_ARIMA and TS_AUTO_ARIMA_OUTLIER; the TS_DOUBLE_ARRAY function provides inputs for TS_GARCH. The other scalar functions return individual scalar result values from the aggregate functions. The supporting scalar functions map the parameters of the TS_GARCH and TS_AUTO_ARIMA functions to the parameters of the C functions contained in the external IMSL libraries.
- Alphabetical List of Functions
This section provides details on all time series functions, including syntax, licensing prerequisites, parameter descriptions, usage, IMSL library mapping, examples, and standards/compatibility information.
- DATASET Example Input Data Table
Time series and forecasting analysis function examples use the following table (called DATASET) as the input data. The DATASET table contains 50 rows of time series data.
Created April 15, 2010. Send feedback on this help topic to Sybase Technical Publications:
pubs@sybase.com