org.apache.spark.mllib.stat.correlation

Interface Correlation

public interface Correlation

Trait for correlation algorithms.

Method Summary

Modifier and Type	Method and Description
double	computeCorrelation(RDD<Object> x, RDD<Object> y) Compute correlation for two datasets.
Matrix	computeCorrelationMatrix(RDD<Vector> X) Compute the correlation matrix S, for the input matrix, where S(i, j) is the correlation between column i and j.
double	computeCorrelationWithMatrixImpl(RDD<Object> x, RDD<Object> y) Combine the two input RDD[Double]s into an RDD[Vector] and compute the correlation using the correlation implementation for RDD[Vector].

Method Detail

computeCorrelation

double computeCorrelation(RDD<Object> x,
                          RDD<Object> y)

Compute correlation for two datasets.

Parameters:

x - (undocumented)

y - (undocumented)

Returns:

(undocumented)

computeCorrelationMatrix

Matrix computeCorrelationMatrix(RDD<Vector> X)

Compute the correlation matrix S, for the input matrix, where S(i, j) is the correlation between column i and j. S(i, j) can be NaN if the correlation is undefined for column i and j.

Parameters:

X - (undocumented)

Returns:

(undocumented)

computeCorrelationWithMatrixImpl

double computeCorrelationWithMatrixImpl(RDD<Object> x,
                                        RDD<Object> y)

Combine the two input RDD[Double]s into an RDD[Vector] and compute the correlation using the correlation implementation for RDD[Vector]. Can be NaN if correlation is undefined for the input vectors.

Parameters:

x - (undocumented)

y - (undocumented)

Returns:

(undocumented)