Java* API Reference for Intel® Data Analytics Acceleration Library 2018 Update 2

List of all members
InitDistributedStep2Local Class Reference

Seconda step of computing initial centroids for the K-Means algorithm on local nodes. More...

Class Constructor

◆ InitDistributedStep2Local() [1/2]

Constructs an algorithm for computing initial centroids for the K-Means algorithm in the first step of the distributed processing mode by copying input objects and parameters of another algorithm

Parameters
contextContext to manage initial centroids for the K-Means algorithm
otherAn algorithm to be used as the source to initialize the input objects and parameters of the algorithm

◆ InitDistributedStep2Local() [2/2]

InitDistributedStep2Local ( DaalContext  context,
Class<? extends Number >  cls,
InitMethod  method,
long  nClusters,
boolean  bFirstIteration 
)

Constructs an algorithm for computing initial centroids for the K-Means algorithm in the first step of the distributed processing mode

Parameters
contextContext to manage initial centroids for the K-Means algorithm
clsData type to use in intermediate computations of initial centroids for the K-Means algorithm, Double.class or Float.class
methodMethod of computing initial centroids for the algorithm, InitMethod
nClustersNumber of initial centroids for the K-Means algorithm
bFirstIterationTrue if step2Local is called for the first time

Detailed Description

References

Member Function Documentation

◆ clone()

Returns the newly allocated algorithm for computing initial centroids for the K-Means algorithm in the first step of the distributed processing mode with a copy of input objects and parameters of this algorithm

Parameters
contextContext to manage initial centroids for the K-Means algorithm
Returns
The newly allocated algorithm

◆ compute()

Computes initial centroids for the K-Means algorithm

Returns
Partial results of computing initial centroids for the K-Means algorithm

◆ finalizeCompute()

InitResult finalizeCompute ( )

Computes the results of K-Means initialization

Returns
Results of K-Means initialization

◆ setPartialResult() [1/2]

void setPartialResult ( InitDistributedStep2LocalPlusPlusPartialResult  partialResult,
boolean  initFlag 
)

Registers user-allocated memory to store partial results of computing initial centroids for the K-Means algorithm

Parameters
partialResultStructure to store partial results of computing initial centroids for the K-Means algorithm
initFlagFlag that specifies initialization of partial results

◆ setPartialResult() [2/2]

void setPartialResult ( InitDistributedStep2LocalPlusPlusPartialResult  partialResult)

Registers user-allocated memory to store partial results of computing initial centroids for the K-Means algorithm

Parameters
partialResultStructure to store partial results of computing initial centroids for the K-Means algorithm

◆ setResult()

void setResult ( InitResult  result)

Registers user-allocated memory to store the results of computing initial centroids for the K-Means algorithm

Parameters
resultStructure to store the results of computing initial centroids for the K-Means algorithm

Member Data Documentation

◆ input

Input data

◆ method

InitMethod method

Method for computing initial centroids

◆ parameter

Parameters for computing initial centroids

◆ partialResult

Partial result of the initialization algorithm


The documentation for this class was generated from the following file:

For more complete information about compiler optimizations, see our Optimization Notice.