Time Series Model Package
TSMSTAT Object
The TSMSTAT object collects model statistics of fit from a TSM object, TSMObj, and stores them in a CAS table. This information is useful for evaluating how well the model fits the dependent series. The CAS table schema that is used for storing the fit statistics is compatible with the schema that the HPFENGINE procedure uses for its OUTSTAT= data set.
Table 17.14 shows the contents of the TSMSTAT object.
Table 17.14: Contents of the TSMSTAT Object
Column | Type | Description |
|---|---|---|
_NAME_ | String | Name of the dependent variable |
_REGION_ | String | Region in which the fit statistics are calculated. You can specify the following values:
|
_MODEL_ | String | Model name |
AADJRSQ | Numeric | Amemiya’s adjusted R-square |
ADJRSQ | Numeric | Adjusted R-square |
AIC | Numeric | Akaike’s information criterion |
AICC | Numeric | Finite sample corrected AIC |
APC | Numeric | Amemiya’s prediction criterion |
DFE | Numeric | Degrees of freedom error |
GMAPE | Numeric | Geometric mean absolute percentage error |
GMAPES | Numeric | Geometric mean absolute error as a percentage of standard deviation |
GMAPPE | Numeric | Geometric mean absolute predictive percentage error |
GMRAE | Numeric | Geometric mean relative absolute error |
GMASPE | Numeric | Geometric mean absolute symmetric percentage error |
MAE | Numeric | Mean absolute error |
MAPE | Numeric | Mean absolute percentage error |
MAPES | Numeric | Mean absolute error as a percentage of standard deviation |
MAPPE | Numeric | Mean absolute predictive percentage error |
MASE | Numeric | Mean absolute scaled error |
MAXAPES | Numeric | Maximum absolute error as a percentage of standard deviation |
MAXERR | Numeric | Maximum error |
MAXPE | Numeric | Maximum percentage error |
MAXPPE | Numeric | Maximum predictive percentage error |
MAXRE | Numeric | Maximum relative error |
MAXSPE | Numeric | Maximum symmetric percentage error |
MDAPE | Numeric | Median absolute percentage error |
MDAPES | Numeric | Median absolute error as a percentage of standard deviation |
MDAPPE | Numeric | Median absolute predictive percentage error |
MDASPE | Numeric | Median absolute symmetric percentage error |
MDRAE | Numeric | Median relative absolute error |
ME | Numeric | Mean error |
MINAPES | Numeric | Minimum absolute error as a percentage of standard deviation |
MINERR | Numeric | Minimum error |
MINPE | Numeric | Minimum percentage error |
MINPPE | Numeric | Minimum predictive percentage error |
MINRE | Numeric | Minimum relative error |
MINSPE | Numeric | Minimum symmetric percentage error |
MPE | Numeric | Mean percentage error |
MPPE | Numeric | Mean predictive percentage error |
MRAE | Numeric | Mean relative absolute error |
MRE | Numeric | Mean relative error |
MSE | Numeric | Mean square error |
MSPE | Numeric | Mean symmetric percentage error |
N | Numeric | Number of observations that were used |
NMISSA | Numeric | Number of missing actual values |
NMISSP | Numeric | Number of missing predicted values |
NOBS | Numeric | Number of observations |
NPARMS | Numeric | Number of parameters |
RMSE | Numeric | Root mean square error |
RSQUARE | Numeric | R-square |
RWRSQ | Numeric | Random walk R-square |
SBC | Numeric | Schwarz Bayesian information criterion |
SMAPE | Numeric | Symmetric mean absolute percentage error |
SSE | Numeric | Sum of square error |
SST | Numeric | Corrected total sum of squares |
TSS | Numeric | Total sum of squares |
UMSE | Numeric | Unbiased mean square error |
URMSE | Numeric | Unbiased root mean square error |
Table 17.15 summarizes the methods that are associated with the TSMSTAT object.
Table 17.15: Methods of the TSMSTAT Object
Figure 17.10 illustrates the data flow through the TSMSTAT object.
Figure 17.10: TSMSTAT Object Data Flow
