![TRUNCNUM function [Numeric]](arrow-left.gif){kind=icon}
![TS_ARMA_CONST function [Time Series]](arrow-right.gif){kind=icon}
This function is available only with RAP – The Trading Edition Enterprise.
Calculates the least-square estimates of parameters for an autoregressive moving average (ARMA) model, and returns the requested autoregressive estimate.
TS_ARMA_AR (timeseries_expression, ar_count, ar_elem, method)
OVER (window-spec)
timeseries_expression A numeric expression, generally a column name, containing an element in a time series.
ar_count An integer containing the number of autoregressive values to compute.
ar_elem An integer identifying the element in the computed AR array that should be returned. ar_elem must be greater than 0 and less than or equal to ar_count.
method (Optional) An integer that identifies the type of procedure used to compute estimates. 0 (the default value) = method of least squares; 1 = method of moments.
window-spec TS_ARMA_AR is an OLAP function requiring an OVER () clause.
TS_ARMA_AR time series function returns a double-precision floating-point value containing the autoregressive estimate. TS_ARMA_AR calls the function imsls_d_arma in the IMSL libraries.
The arguments of TS_ARMA_AR map to the IMSL library function imsls_d_arma as follows:
params = imsls_d_arma(n_objs, z, p, q, methodID, 0);
n_objs Contains the number of rows in the current window frame.
z[] Contains the value of timeseries_expression for the current window frame.
p Maps to the user-defined aggregate function argument ar_count.
methodID Maps to the method argument of TS_ARMA_AR. Can be set to either IMSLS_METHOD_OF_MOMENTS or IMSLS_LEAST_SQUARES.
For detailed information on how imsls_d_arma performs time series calculations, see IMSL Numerical Library User’s Guide: Volume 2 of 2 C Stat Library.
This example shows an input data table, a SQL statement containing the TS_ARMA_AR function, and the data values returned by the function. This example uses the following table (called DATASET) as its input data. The DATASET table contains 50 rows of time series data.
rownum |
data |
|---|---|
1 |
0.315523 |
2 |
0.485859 |
3 |
0.676886 |
4 |
1.97381 |
5 |
2.77555 |
6 |
2.73657 |
7 |
2.64233 |
8 |
4.26118 |
9 |
3.13641 |
10 |
4.16566 |
11 |
2.95952 |
12 |
2.14504 |
13 |
1.98799 |
14 |
0.805859 |
15 |
0.833405 |
16 |
2.29075 |
17 |
1.30045 |
18 |
0.467122 |
19 |
-0.170107 |
20 |
-0.256657 |
21 |
-0.382597 |
22 |
-0.505511 |
23 |
-1.90147 |
24 |
-0.981688 |
25 |
-1.43116 |
26 |
-1.39389 |
27 |
-2.34823 |
28 |
-2.91122 |
29 |
-0.927423 |
30 |
-0.044383 |
31 |
-0.389648 |
32 |
0.545008 |
33 |
0.614096 |
34 |
0.364668 |
35 |
1.16043 |
36 |
-0.654063 |
37 |
0.616094 |
38 |
2.00875 |
39 |
1.86696 |
40 |
2.80171 |
41 |
3.78422 |
42 |
4.11499 |
43 |
2.77188 |
44 |
4.00312 |
45 |
4.21298 |
46 |
5.00413 |
47 |
4.74498 |
48 |
4.89621 |
49 |
3.93273 |
50 |
4.31592 |
The following SQL statement returns the first element of an autoregressive estimate consisting of one value from the data column using the method of least squares:
SELECT TS_ARMA_AR(data,1,1,0) OVER (ORDER BY rownum ROWS BETWEEN UNBOUNDED PRECEDING AND UNBOUNDED FOLLOWING) AS res FROM DATASET
Sybase IQ returns 50 rows, each containing the same value:
res |
|---|
0.898793 |
0.898793 |
0.898793 |
0.898793 |
0.898793 |
0.898793 |
0.898793 |
0.898793 |
0.898793 |
... |
0.898793 |
Chapter 2, “Using OLAP” in the System Administration Guide: Volume 2
IMSL Numerical Library User’s Guide: Volume 2 of 2 C Stat Library
