SECTION 15.22 INVERSE CUMULATIVE NORMAL
The Subroutine NRMCDFI computes the inverse of the cumulative normal distribution function. Denoting by F(x;mu,var) the value at x of the cumulative normal distribution function with mean given by mu and variance given by var, NRMCDFI computes for a given number y the value x for which
x=F-1(y)
Be sure to CALL DFLT before calling NRMCDFI and to include the COMMON/BPRINT/ common block. The call is
CALL NRMCDFI(XMU,VAR,Y,ACC,ITERL,SOL,IER) where XMU = double precision mean (input) VAR = double precision variance (input) Y = double precision value of the CDF (input) (must be between 5.55111D-17 and 1.-5.55111D-17) ACC = double precision accuracy required (e.g.,0.00001) (input) ITERL = integer iteration limit (input) SOL = double precision solution (output) IER = error return (=-3 input error, =-1 iteration limit exceeded) (output)