DATASET subcommand (SPATIAL TEMPORAL PREDICTION command)

The DATASET subcommand identifies the data sources that are used in the analysis. The datasets are defined in a map specification file that is created by the SPATIAL MAPSPEC command. You can specify a maximum of two DATASET subcommands.

The following requirements apply when there are two data sources:

Kernel density estimates (KDE) are calculated and used as the target.
The map specification file must include a map (SPATIAL MAPSPEC command, MAP subcommand).
Both data sources must contain a time field.
One of the data sources must specify ROLE=EVENTDENSITY on the DATA subcommand of the SPATIAL MAPSPEC command. This data source must contain coordinates.
The other data source must specify ROLE=CONTEXT on the DATA subcommand of the SPATIAL MAPSPEC command. This data source must be matched to the map with key values.

DATAID=id

The IDENTIFIER value of the dataset, which is specified on a DATA subcommand of the SPATIAL MAPSPEC command.

KDENAME=rootname

Root field name for calculated kernel density estimates. If a field name with that name exists, a unique new name is generated by appending an integer value to the name.

TARGET=fieldname

The name of the target field.

If there is only one DATASET subcommand, a single target field is required.
If there are two DATASET subcommands, the TARGET keyword is not valid.

INPUTS=fieldlist

The inputs (predictors) in the dataset. Inputs for each dataset are specified on separate DATASET subcommands.

TIMEINDEX=fieldlist

The fields that define time periods. Multiple fields can be specified to define periodicity. Each DATASET subcommand must contain a TIMEINDEX specification, and they must define the same time interval.

For the specified fields, overrides the aggregation method specified on the AGGREGATION subcommand and uses the specified method.

The method name is enclosed in parentheses.
Specifications for multiple fields are separated by semi-colons.