Vector32 Example Using TD_KMeans - Teradata VantageCloud Lake

Lake - Analyze Your Data with ClearScape Analytics™

Deployment
VantageCloud
Edition
Lake
Product
Teradata VantageCloud Lake
Release Number
Published
February 2025
ft:locale
en-US
ft:lastEdition
2026-02-20
dita:mapPath
tcl1683670667798.ditamap
dita:ditavalPath
pny1626732985837.ditaval
dita:id
tcl1683670667798

Input

DROP TABLE input_table_udt;
CREATE MULTISET TABLE input_table_udt(id INTEGER, array_col Vector32) PRIMARY INDEX ( id );
INSERT INTO input_table_udt values(1, '-0.06944,0.080352,0.045963,0.006985,-0.000496');
INSERT INTO input_table_udt values(2, '-0.025461,0.045302,-0.028107,0.031248,0.00077');
INSERT INTO input_table_udt values(3, '-0.028107,0.016781,-0.023147,-0.068779,-0.07936');
INSERT INTO input_table_udt values(4, '-0.030091,0.027611,-0.047616,-0.025461,0.029595');

input_table_udt

id array_col
1 -0.06944,0.080352,0.045963,0.006985,-0.000496
2 -0.025461,0.045302,-0.028107,0.031248,0.00077
3 -0.028107,0.016781,-0.023147,-0.068779,-0.07936
4 -0.030091,0.027611,-0.047616,-0.025461,0.029595
DROP TBALE initial_centroids_table_udt;
CREATE MULTISET TABLE initial_centroids_table_udt AS (SELECT * FROM input_table_udt WHERE id in (2,4)) with data;

Query (InitialCentroidsTable not provided)

SELECT td_clusterid_kmeans, cast(array_col as Varchar(100)) AS array_col, td_size_kmeans, td_withinss_kmeans, id, td_modelinfo_kmeans
FROM TD_KMeans (
ON input_table_udt as InputTable
USING
IdColumn('id')
TargetColumns('array_col')
NumClusters(2)
Seed(0)
StopThreshold(0.0395)
MaxIterNum(3)
EmbeddingSize(5)
UseSIMD('t')  
)AS dt;

Output

td_cluster_kmeans array_col td_ksize_kmeans td_withinss_kmeans id td_modelinfo_kmeans
0 -0.06944,0.080352,0.045963,0.006985,-0.000496 1 0.00000000000000E 000 NULL NULL
1 -0.0278863333333333,0.029898,-0.0329566666666667,-0.0209973333333333,-0.0163316666666667 3 1.21668890066667E-002 NULL NULL
NULL NULL NULL NULL NULL Converged : True
NULL NULL NULL NULL NULL Number of Iterations : 2
NULL NULL NULL NULL NULL Number of Clusters : 2
NULL NULL NULL NULL NULL Total_WithinSS : 1.21668890066667E-02
NULL NULL NULL NULL NULL Between_SS : 8.65080482533333E-03
NULL NULL NULL NULL NULL Method for InitialCentroids : Random

Query (InitialCentroidsTable is provided)

SELECT td_clusterid_kmeans, cast(array_col as Varchar(100)) AS array_col, td_size_kmeans, td_withinss_kmeans, id, td_modelinfo_kmeans FROM TD_KMeans (
ON input_table_udt as InputTable
ON initial_centroids_table_udt as InitialCentroidsTable DIMENSION
USING
IdColumn('id')
TargetColumns('array_col')
StopThreshold(0.0395)
MaxIterNum(3)
EmbeddingSize(5)
UseSIMD('t')  
)AS dt;

Output

td_cluster_kmeans array_col td_ksize_kmeans td_withinss_kmeans id td_modelinfo_kmeans
0 -0.0474505,0.062827,0.008928,0.0191165,0.000137 2 4.61965788300000E-003 NULL NULL
1 -0.029099,0.022196,-0.0353815,-0.04712,-0.0248825 2 7.23379913300000E-003 NULL NULL
NULL NULL NULL NULL NULL Converged : True
NULL NULL NULL NULL NULL Number of Iterations : 2
NULL NULL NULL NULL NULL Number of Clusters : 2
NULL NULL NULL NULL NULL Total_WithinSS : 1.18534570160000E-02
NULL NULL NULL NULL NULL Between_SS : 8.96423681600000E-03
NULL NULL NULL NULL NULL Method for InitialCentroids : Externally supplied InitialCentroidsTable

Query (OutputClusterAssignment is set to true)

SELECT * FROM TD_KMeans (
ON input_table_udt AS InputTable
ON initial_centroids_table_udt AS InitialCentroidsTable DIMENSION
USING
IdColumn('id')
TargetColumns('array_col')
StopThreshold(0.0395)
MaxIterNum(3)
OutputClusterAssignment('true')
EmbeddingSize(5)
UseSIMD('t')  
)AS dt;

Output

id td_clusterid_kmeans
1 0
2 0
3 1
4 1

Query (OutputClusterAssignment is set to true, and ModelTable (OUT TABLE clause) is provided)

SELECT * FROM TD_KMeans (
ON input_table_udt AS InputTable
ON initial_centroids_table_udt AS InitialCentroidsTable DIMENSION
OUT TABLE ModelTable(kmeans_model_udt)
USING
IdColumn('id')
TargetColumns('array_col')
StopThreshold(0.0395)
MaxIterNum(3)
OutputClusterAssignment('true')
EmbeddingSize(5)
UseSIMD('t')  
)AS dt;

Output

id td_clusterid_kmeans
1 0
2 0
3 1
4 1

ModelTable: kmeans_model

td_cluster_kmeans array_col td_ksize_kmeans td_withinss_kmeans id td_modelinfo_kmeans
0 -0.0474505,0.062827,0.008928,0.0191165,0.000137 2 4.61965788300000E-003 NULL NULL
1 -0.029099,0.022196,-0.0353815,-0.04712,-0.0248825 2 7.23379913300000E-003 NULL NULL
NULL NULL NULL NULL NULL Converged : True
NULL NULL NULL NULL NULL Number of Iterations : 2
NULL NULL NULL NULL NULL Number of Clusters : 2
NULL NULL NULL NULL NULL Total_WithinSS : 1.18534570160000E-02
NULL NULL NULL NULL NULL Between_SS : 8.96423681600000E-03
NULL NULL NULL NULL NULL Method for InitialCentroids : Externally supplied InitialCentroidsTable