Probabilities.FCDFInv Method

Overload List

#SignatureDescription
1procedure FCDFInv(const p: TDenseMtxVec; Nu1: Integer; Nu2: Integer; const Res: TDenseMtxVec);
2procedure FCDFInv(p: Double; const Nu1: TMtxVecInt; const Nu2: TMtxVecInt; const Res: TDenseMtxVec);Fisher(F) PPF (vectorized).
3function FCDFInv(p: Double; Nu1: Integer; Nu2: Integer): Double;F distribution inverse CDF (quantile / point percent function, PPF).

Overload 1: procedure FCDFInv(const p: TDenseMtxVec; Nu1: Integer; Nu2: Integer; const Res: TDenseMtxVec);

#NameTypeDescription
1pTDenseMtxVec
2Nu1Integer
3Nu2Integer
4ResTDenseMtxVec

Result: stored in self (calling object)

Overload 2: procedure FCDFInv(p: Double; const Nu1: TMtxVecInt; const Nu2: TMtxVecInt; const Res: TDenseMtxVec);

Fisher(F) PPF (vectorized).

#NameDescription
1pDefines distribution probabilities, real vector or matrix with values within closed interval [0,1].
2Nu1Defines distribution degrees of freedom. Nu1 must be a positive integer.
3Nu2Defines distribution degrees of freedom. Nu2 must be a positive integer.
4ResAfter calculation stores the PPF calculated from P, Nu1 and Nu2. Length and Complex properties of Res are adjusted automatically to match Length and Complex properties of P.

Result: stored in self (calling object)

Overload 3: function FCDFInv(p: Double; Nu1: Integer; Nu2: Integer): Double;

F distribution inverse CDF (quantile / point percent function, PPF).

#NameDescription
1pProbability, real value on the closed interval [0,1].
2Nu1Numerator degrees of freedom, integer > 0.
3Nu2Denominator degrees of freedom, integer > 0.

Returns: Double - the value x with FCDF(x,Nu1,Nu2)=p for degrees of freedom Nu1 and Nu2. Returns NAN when Nu1 <= 0 or Nu2 <= 0; returns +INF at p=1.

Remarks:

Computes the inverse of the F CDF (the quantile function)

inverse CDF(p|nu _1, nu _2) = F^(-1)(p|nu _1, nu _2 ) , p(x)=CDF(x|nu _1, nu _2)

via the beta quantile z= BetaCDFInv(1-p, Nu2/2, Nu1/2) and x=(nu_2/z-nu_2)/nu_1. Domain: p in [0,1], integer nu_1,nu_2 > 0. Behaviour: returns +INF at p=1; for nu_1 <= 0 or nu_2 <= 0 returns NAN.

See Also: Probabilities.FPDF, Probabilities.FCDF