22 import daal.algorithms.kmeans.init
23 from daal.algorithms
import kmeans
25 utils_folder = os.path.realpath(os.path.abspath(os.path.dirname(os.path.dirname(__file__))))
26 if utils_folder
not in sys.path:
27 sys.path.insert(0, utils_folder)
28 from utils
import printNumericTable, createSparseTable
30 DAAL_PREFIX = os.path.join(
'..',
'data')
33 datasetFileName = os.path.join(DAAL_PREFIX,
'batch',
'kmeans_csr.csv')
39 if __name__ ==
"__main__":
42 dataTable = createSparseTable(datasetFileName)
45 init = kmeans.init.Batch(nClusters, method=kmeans.init.randomDense)
47 init.input.set(kmeans.init.data, dataTable)
50 centroids = res.get(kmeans.init.centroids)
53 algorithm = kmeans.Batch(nClusters, nIterations, method=kmeans.lloydCSR)
55 algorithm.input.set(kmeans.data, dataTable)
56 algorithm.input.set(kmeans.inputCentroids, centroids)
58 res = algorithm.compute()
61 printNumericTable(res.get(kmeans.assignments),
"First 10 cluster assignments:", 10)
62 printNumericTable(res.get(kmeans.centroids),
"First 10 dimensions of centroids:", 20, 10)
63 printNumericTable(res.get(kmeans.objectiveFunction),
"Objective function value:")