CRITERIA Subcommand

The CRITERIA subcommand includes control parameters to set the normalization, significance level, and permutation procedure.

NORMALIZATION

The NORMALIZATION subcommand controls the normalization method applied to the variables in the analysis. See the following available options:

Option	Description
NONE	No normalization is applied. This option uses the raw values of the variables.
MINMAX	Normalizes the data by using min-max scaling. Each variable is scaled to a range between 0 and 1, based on its minimum and maximum values.
ZSCORE	Normalizes the data by converting each variable to its z-score. This transforms the data to have a mean of 0 and a standard deviation of 1.
ROBUST	A robust normalization method that uses the median and interquartile range (IQR) for scaling. This option is more resistant to outliers compared to the min-max or z-score methods
LOG	Applies a logarithmic transformation to the variables. This can be useful when the data has a skewed distribution, as it can help normalize the data.

CILEVEL

The CILEVEL subcommand specifies the significance level that is used for computing credible intervals. The value must be a single double value within the range (0, 100), with 95 being the default setting.

PERMUTATION

The PERMUTATION subcommand specifies the settings for the permutation procedure that is used to assess the statistical significance of the distance correlation. Permutation testing is a nonparametric method that involves repeatedly reshuffling the data and calculating the correlation for each permutation. The results are then compared to the observed distance correlation to assess its significance. This helps determine whether the observed correlation is likely to occur by chance.

Option	Description
VARIABLE={varlist}:	This option allows the user to specify the variables for which the permutation procedure is applied. varlist is a list of variables whose distance correlation is evaluated for significance.
SEED={n}	The SEED parameter specifies the seed value for the random number generator, ensuring the reproducibility of the permutation procedure. By default, a random seed is used, but the user can specify a specific value to control the randomness.
MAXITERATIONS={100**1, value}	This option controls the maximum number of iterations (or permutations) for the test. The default value is 100, which means the data is reshuffled and correlations are calculated up to 100 times. The user can specify a different number of iterations if wanted, which might increase the precision of the test but also requires more computational time.
SIGNIFICANCELEVEL={0.05**, value}	This parameter sets the significance level (alpha level) for the permutation test. The default significance level is 0.05, meaning that a result is considered statistically significant if the observed correlation is in the top 5% of permutations. The user can adjust this level by specifying a different value, depending on the desired strictness of the test.

/CRITERIA  NORMALIZATION(MINMAX)  CILEVEL(95)  PERMUTATION(SEED=2000000 MAXITERATIONS=100 
    SIGNIFICANCELEVEL=0.05)