The PhysicalData (PD) object contains the physical description of the data used by the mining processes. It usually refers to a table in a relational database. It contains the information about the location of the data (csv file, database table etc.) and the physical data type (integer/double/string/notSpecified). It also contains some basic statistics (nulls, distinct etc.). The physical data object may be used as a data source (for example in a model building task) or as a data output (in a model application task).
Table 12.4. Connected objects
| Status | Name | Description | Type | Reference |
|---|---|---|---|---|
| included | alias | stores the alias of the database which contains the selected data | Database Alias | own |
| included | attributes | contains a series of statistics, separate for each variable; the statistics are computed only when other tasks are executed, such as CalculateStatiscticsTask, BuildingModel | Elements | own |
| included | statistics | this object is responsible for storing the statistics | Statistics | own |