Table | Description |
---|---|
InputTable | Contains features by which to cluster the data. |
CentroidsTable | [Optional] Contains initial seed means for clusters. Has same schema as KMeans Output OutputTable, which depends on UnpackColumns syntax element. |
InputTable Schema
It is important to normalize the input variables before calling this function. For details, see Normalized Input.
Column | Data Type | Description |
---|---|---|
id | INTEGER | Identifier of user or item. |
dimension_i | DOUBLE PRECISION | [Column appears once for each dimension.] Data in dimension i. Each dimension is a feature by which to cluster data. The number of dimensions is implicitly derived from the data. For example, if the application clusters points on the surface of the earth by latitude and longitude, the input table has three columns: point identifier, latitude, and longitude. The function clusters the data in the latitude and longitude columns; that is, the data has two dimensions. |