XProfan

English

p.specht

wants one Mittelwerte different great Stichproben on Plausibilität vergleichen, must one üblicherweise Tabellenwerke zurate wander. The t-distribution is the Gauss-schen Glockenkurve ("Normalverteilung") of course similar, but flacher. tappt im dunkeln goes first ex Stichprobengrößen explicit over 30 slow in these over (details see Wikipedia).
whom 4 To 5 Nachkommastellen accuracy wealthy (übliche Tabellengenauigkeit), the is with the TOMS-Algorithm 344 the ACM well bedient, presupposed it holds itself on the copyright the ACM. The rendition here results anyway solely To Demonstrationszwecken and without jedwede Gewähr!

Compile Mark Separation

Window Title "Proc Student_TTEST:  t-distribution for p<=0.5 with Probengröße 3"
CLS
print Student_TTEST(0.5, 3)
waitinput

Proc Student_TTEST :parameters T!,DF&

    ' AUSZUG AUS ACM TOMS Algorithm 344 of John Burkardt
    ' Reference:
    '     Milton Abramowitz and Irene Stegun,
    '     Handbook of Mathematical Functions,
    '     US Department of Commerce, 1964.
    '
    '     Stephen Wolfram,
    '     The Mathematica Book,
    '     Fourth Edition,
    '     Wolfram Media / cambridge University Press, 1999.
    Declare D1!,D2!,F1!,F2!,I&,N&,T1!,T2!,ans!
    D1!=0.63661977236758134'= 2 / PI
    case DF&<1:RETURN -99999999
    ' BEGIN COMPUTATION OF SERIES
    T2!=sqr(T!)/DF&:T1!=SQRT(T2!):T2!=1/(1+T2!):casenot DF& mod 2: GOTO "G5_"
    ' DF IS AN ODD INTEGER:
    ANS! = 1-D1!*ArcTAN(T1!)
    case DF&=1:GOTO "G4_"
    D2!=D1!*T1!*T2!:ANS!=ANS!-D2!
    case DF&=3:GOTO "G4_"
    F1!=0
    G2_:
    N&=(DF&-2)\2

    whileloop n&:i&=&Loop

        F2!=2*I&-F1!
        D2!=D2!*T2!*F2!/(F2!+1)
        ANS!=ANS!-D2!

    endwhile

    G4_:
    ' COMMON RETURN AFTER COMPUTATION
    case ANS!<0:ANS!=0
    RETURN ANS!
    G5_:
    ' DF IS AN EVEN INTEGER
    D2!=T1!*SQRT(T2!)
    ANS!=1-D2!
    case DF&=2:GOTO "G4_"
    F1!=1:GOTO "G2_"

ENDPROC