BUILDOPTIONS subcommand (TCM MODEL command)

Specifies the estimation period that defines the set of observations that are used for the analysis.

ALL

Specifies that all observations are used. This setting is the default. When ALL is specified, the estimation period starts at the time of the earliest observation and ends at the time of the latest observation across all series.

TIMES

Specifies a range of observations by start and end time.

If START is specified without END, then all observations from START to the last observation (across all series) are used. If END is specified without START, then all observations from the first observation (across all series) through END are used.
The FORMAT keyword specifies the format of the values for START and END. It is required when TIMETYPE=DATETIMEVAR or TIMETYPE=ISO8601, and is ignored otherwise. Values for START and END must be quoted in this case.
VARIABLE

Specifies to use the variable format of the field that is specified on the TIMEVAR keyword; for example, mm/dd/yyyy. This option only applies when TIMETYPE=DATETIMEVAR.

ISO8601
Specifies that values for START and END are in the general form of yyyy-mm-dd hh:mm:ss. FORMAT=ISO8601 must be specified when TIMETYPE=ISO8601.
- Dates are specified in the form yyyy-mm-dd. Datetimes are specified in the form yyyy-mm-dd hh:mm:ss. Times are specified in the form hh:mm:ss.
- Two digits are not required for month, day, hours, minutes or seconds specifications. For example, 2014-1-1 is interpreted as 2014-01-01 and 3:2:1 is interpreted as 03:02:01.
- Values can be specified with reduced precision. For example, 2014-10 is interpreted as 2014-10-01 and 03:02 is interpreted as 03:02:00.
For TIMETYPE=CYCLICPERIODS, START and END are lists of integers, consisting of one value for each field that is specified by the TIMEVAR keyword. The first value is the value for the highest level cycle field, followed by the next highest cycle field, and so on.
For example, the following syntax specifies the start and end time when there are two cycle levels:
```
START = 3 6 2  END = 5 1 4
```
The start of the estimation period is defined by a value of 3 for the second-level cycle, a value of 6 for the first-level cycle, and a value of 2 for the period field. The end of the estimation period is defined by a value of 5 for the second- level cycle, a value of 1 for the first-level cycle, and a value of 4 for the period field.
For TIMETYPE=NONE, START and END consist of a single value that specifies the starting record number and the ending record number. For example, the following syntax specifies an estimation period that starts at record 1 and ends at record 60:
```
START = 1  END = 60
```
If a date specification exists for the active dataset, then START and END are lists of integers that consist of one value for each of the components of the date specification. For example, if the date specification is YEAR and MONTH, then the lists consist of two items, one for the year and one for the month. The order of the values is from the highest level component to the lowest level component. For example, the following syntax specifies start and end time for a date specification that is defined by YEAR and MONTH:
```
START = 2008 1  END = 2013 12
```
The start of the estimation period is defined by a value of 2008 for the year and 1 for the month. The end of the estimation period is defined by a value of 2013 for the year and 12 for the month.

OBSERVATIONS

Defines the estimation period as a specified number of time intervals that start at the earliest time interval or end at the latest time interval in the data, with an optional offset. In this context, the time interval refers to the time interval of the analysis.

TYPE: TYPE=NEWEST specifies to use the most recent time intervals. TYPE=OLDEST specifies to use the earliest time intervals.
OFFSET: Specifies an offset from the earliest time interval (when TYPE=OLDEST) or the most recent time interval (when TYPE=NEWEST). The value of OFFSET must be an integer greater than or equal to 0. The default is 0 and specifies to start at the earliest time interval or end at the most recent time interval.
NUMBER: Specifies the number of time intervals to include. The value must be an integer that is greater than or equal to 2. The default is 1000. If the specified value exceeds the number of time intervals in the data, then all time intervals, starting from the optional offset, are included.

Specifies the number of time intervals to forecast beyond the end of the estimation period. The time interval in this case is the time interval of the analysis, which is specified by the ANALYSISINTERVAL keyword. The default value of FORECASTPERIODS is 0, which means that forecasts are not created.

Specifies the confidence level for computing confidence intervals for forecasts and parameter estimates. The value must be positive number that is greater than 0 and less than 100. The default is 95.

The MAXINPUTS keyword specifies the maximum number of inputs that can be included in the model for any target.

The minimum value is 1 and the maximum value is 20.
The default value, which is specified by DEFAULT, is the lesser of 5 or the total number of series.
The model for each target always includes lagged values of itself, so setting this value to 1 specifies that the only input is the target itself.

Specifies the number of lag terms that are included for each time series in the system. The minimum value is 1 and the maximum value is 20. The default is AUTO, which automatically determines the number of lags from the time interval that is used for the analysis. For example, if the time interval is months (with an increment of one month) then the number of lags is 12.

This setting controls the iterative process that is used for determining the best set of inputs for each target. You can specify any value that is greater than zero. The default is 0.001. Model tolerance is a stop criterion for predictor selection. It can affect the number of predictors that are included in the final model. But if a target can predict itself very well, other predictors may not be included in the final model. Some trial and error may be required (for example, if you have this value set to high, you can try setting it to a smaller value to see if other predictors can be included or not).

An observation is flagged as an outlier if the probability, as calculated from the model, that it is an outlier exceeds this threshold. You can specify a value in the range 50 - 100. The default value is 95.