CALL td_analyze (
'kmeans',
'required_parameter_list [ optional_parameter; [...] ]'
);
- required_parameter_list
database = input_database_name;
tablename = input_table_name;
columns = { column_name [,...] | keyword };
outputdatabase = output_database_name;
outputtablename = output_table_name;
kvalue = k_value;
- optional_parameter
{ columnstoexclude = column_name [,...] |
continuation = { true | false } |
iterations = iterations |
operatordatabase = operator_database_name |
overwrite = { true | false } |
threshold = threshold
}
Syntax Elements
- database
- The database containing the input table.
- tablename
- The name of the table containing the data to cluster.
- columns
- The columns to analyze.
keyword |
Description |
all |
All columns. |
allnumeric |
All numeric columns. |
- outputdatabase
- The database to contain the resulting output table that represents a cluster model.
- outputtablename
- The name of the output table representing the cluster model.
- kvalue
- The number of clusters to be contained in the cluster model.
- columnstoexclude
- [Optional] The columns to exclude when columns specifies a keyword.
- continuation
- [Optional] Whether clustering begins with values determined by pre-existing result tables rather than random values.
- Default: false
- iterations
- [Optional] The maximum number of iterations to perform during modeling.
- Default: 50
- operatordatabase
- [Optional] The database where the table operators that td_analyze calls reside.
- Default behavior: The function searches the standard search path for table operators.
- overwrite
- [Optional] Whether to drop the output tables before creating new ones.
- Default: true
- threshold
- [Optional] The decimal value that determines if the algorithm has converged, based on how much the cluster centroids change from one iteration to the next.
- Default: .001