TARGETFILTER subcommand (TCM MODEL command)
The TARGETFILTER
subcommand specifies a set of target series for which
output is displayed. The content of the output for the specified series is determined by the
FILTEREDOUTPUT
subcommand. You can specify particular series, for which
output is displayed, from the SERIESFILTER
subcommand.
- BESTFIT
- Specifies to generate output for the target series that have the best-fitting models.
By default, output is generated for the top 10 models according to the value of the root
mean square percentage error.
- The
N
keyword specifies the number of models to include, as an integer. If the count exceeds the number of models, all models are shown. The default is 10. - The
PCT
keyword specifies the number of models to include, as a percentage of the total number of models. The percentage value must be an integer between 1 and 100. - The
FITSTAT
keyword specifies the fit statistic that is used to determine the best-fitting models. The default isRSQUARE
.- RSQUARE
- Goodness-of-fit measure of a linear model, sometimes called the coefficient of determination. It is the proportion of variation in the target variable explained by the model. It ranges in value from 0 to 1. Small values indicate that the model does not fit the data well.
- RMSPE
- Root Mean Square Percentage Error. A measure of how much the model-predicted values differ from the observed values of the series. It is independent of the units that are used and can therefore be used to compare series with different units.
- RMSE
- Root Mean Square Error. The square root of mean square error. A measure of how much a dependent series varies from its model-predicted level, expressed in the same units as the dependent series.
- BIC
- Bayesian Information Criterion. A measure for selecting and comparing models based on the -2 reduced log likelihood. Smaller values indicate better models. The BIC also "penalizes" overparameterized models (complex models with a large number of inputs, for example), but more strictly than the AIC.
- AIC
- Akaike Information Criterion. A measure for selecting and comparing models based on the -2 reduced log likelihood. Smaller values indicate better models. The AIC "penalizes" overparameterized models (complex models with a large number of inputs, for example).
- The
- ALLTARGETS
- Specifies that output is generated for all target series.
- TARGETSNOINPUTS
- Specifies that only target series with no inputs (other than lagged values of the target itself) are included.