DISTANCE CORRELATION

DISTANCE CORRELATION is available in Statistics Base Edition

Run the DISTANCE CORRELATION procedure to assess the dependence between two random variables or multivariate data, and capture the nonlinear relationships. It is calculated by dividing the distance covariance with the product of distance variances of the variables.

Note:

Replace variable list with appropriate variable names.
** denotes default options.

DISTANCE CORRELATION VARIABLE = variable list
[ /ID
   { VARIABLE = varname } ]
[ /CRITERIA
  [ NORMALIZATION ({NONE^**, MINMAX, ZSCORE, ROBUST, LOG}) ]
  [ CILEVEL ({95^**, value}) ]
  [ PERMUTATION
    [ (SEED = {n}
    MAXITERATIONS = {100^**, value}
    SIGNIFICANCELEVEL = {0.05^**, value}) ]
]

[ /PRINT
  [ DETAILS ]
  [ DISTCORRCOEFF ]
  [ CORREST ]
  [ DISTVAR ]
  [ DISTMATRIX
    [ CONFIGURE = {varlist} ]
    [ SAVE= {TRUE, FALSE^**} ]
    [ OUTFILE = {filename} ]
]

[ /PLOT
  [ SCATTERPLOT [(XAXIS = {varname^**}
     YAXIS = {varname^**} ) ]
]