ERF ERF returns a real floating-point value that is an approximation of the error function of x. ERF( x) x Real expression. The result has the base and precision of x, and a value given by: