Java* API Reference for Intel® Data Analytics Acceleration Library 2018 Update 1

List of all members
InitDistributedStep1Local Class Reference

First step of computing initial clusters for the K-Means algorithm on local nodes. More...

Class Constructor

Constructs an algorithm for computing initial clusters 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 clusters for the K-Means algorithm
otherAn algorithm to be used as the source to initialize the input objects and parameters of the algorithm
InitDistributedStep1Local ( DaalContext  context,
Class<?extends Number >  cls,
InitMethod  method,
long  nClusters,
long  nRowsTotal,
long  offset 
)

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

Parameters
contextContext to manage initial clusters for the K-Means algorithm
clsData type to use in intermediate computations of initial clusters for the K-Means algorithm, Double.class or Float.class
methodMethod of computing initial clusters for the algorithm, InitMethod
nClustersNumber of initial clusters for the K-Means algorithm
nRowsTotalNumber of rows in all data sets in the distributed processing mode
offsetOffset in the total data set specifying the start of a block stored on a given local node

Detailed Description

References

Member Function Documentation

Returns the newly allocated algorithm for computing initial clusters 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 clusters for the K-Means algorithm
Returns
The newly allocated algorithm
InitPartialResult compute ( )

Computes initial clusters for the K-Means algorithm

Returns
Partial results of computing initial clusters for the K-Means algorithm
InitResult finalizeCompute ( )

Computes the results of K-Means initialization

Returns
Results of K-Means initialization
void setPartialResult ( InitPartialResult  partialResult,
boolean  initFlag 
)

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

Parameters
partialResultStructure to store partial results of computing initial clusters for the K-Means algorithm
initFlagFlag that specifies initialization of partial results
void setPartialResult ( InitPartialResult  partialResult)

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

Parameters
partialResultStructure to store partial results of computing initial clusters for the K-Means algorithm
void setResult ( InitResult  result)

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

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

Member Data Documentation

Input data

InitMethod method

Method for computing initial clusters

InitParameter parameter

Parameters for computing initial clusters

InitPartialResult partialResult
protected

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.