FisherTransform

Declaration:

FisherTransform (x: double): double;

The function FisherTransform returns an approximately normally distributed value rather than a skewed one. The parameter x theoretically has to be in the range between -1 and +1. Under practical circumstances - due to the limited resolution of floating point numbers - the valid range of x is between -0.9999999999999999 and +0.9999999999999999, values outside this range result in an exact value of +/-20 (depending on the sign of x).

Hint:

The Fisher transformation is typically used for hypothesis testing on the correlation coefficient.