!*********************************************************
!*      Program to demonstrate Chi-square Statistic      *
!* ----------------------------------------------------- *
!*   Reference: BASIC Scientific Subroutines, Vol. II    *
!*   By F.R. Ruckdeschel, BYTE/McGRAWW-HILL, 1981 [1].   *
!*                                                       *
!*                     F90 version by J-P Moreau, Paris  *
!*                            (www.jpmoreau.fr)          *
!* ----------------------------------------------------- *
!* SAMPLE RUN:                                           *
!*                                                       *
!*  CHI-SQUARE CUMULATIVE DISTRIBUTION APPROXIMATION     *
!*                                                       *
!*   P(X)        X2       REAL X2                        *
!*  -----------------------------                        *
!*  0.05     124.5347     124.3                          *
!*  0.10     118.6326     118.5                          *
!*  0.15     114.5999                                    *
!*  0.20     111.4342                                    *
!*  0.25     108.7913     109.1                          *
!*  0.30     106.5055                                    *
!*  0.35     104.4821                                    *
!*  0.40     102.6610                                    *
!*  0.45     101.0018                                    *
!*  0.50      99.1944      99.3                          *
!*  0.55      97.6863                                    *
!*  0.60      96.0812                                    *
!*  0.65      94.3595                                    *
!*  0.70      92.4934                                    *
!*  0.75      90.4428      90.1                          *
!*  0.80      88.1446                                    *
!*  0.85      85.4893                                    *
!*  0.90      82.2522      82.4                          *
!*  0.95      77.7867      77.9                          *
!*  1.00       0.0000                                    *
!*  -----------------------------                        *
!*                                                       *
!*********************************************************
PROGRAM CHISQA

  real*8  x,y
  integer i,m;
  character*5 x1(20)

  do i = 1, 20  
    x1(i) = ' '
  end do
  x1(1)  = '124.3'; x1(2)  = '118.5';
  x1(5)  = '109.1'; x1(10) = ' 99.3';
  x1(15) = ' 90.1'; x1(18) = ' 82.4'; x1(19) = ' 77.9';
  print *,'CHI-SQUARE CUMULATIVE DISTRIBUTION APPROXIMATION'
  print *,' P(X)        X2       REAL X2 '
  print *,'-----------------------------'
  m = 100; i = 0
  y = 0.05
  do while (y < 1.01)
    i = i + 1
    call Chi_Square(m,x,y)
    write(*,25) y,x,x1(i)
    y = y + 0.05
  end do
  print *,'-----------------------------'
  stop
  
25 format(F5.2,'     ',F8.4,'     ',A5)   
end

!*****************************************************
!* Chi-square cumulative distribution approximation  *
!* Good for m > 100.                                 *
!* Reference: Statistics Manual, Crow, Maxfield and  *
!* Davis (Dover, 1960).                              *
!* The input value is y, the probability, the output * 
!* value is the corresponding Chi-square statistic.  *
!*****************************************************
subroutine Chi_Square(m,x,y)  
  !Labels: 100, 200, 210
  real*8 x,y,z
  x = y
  !Guard against zero discontinuity
  if (x <= 0) x = dexp(-100.d0)
  if (x > 0.5) goto 100
  x = -dlog(x)
  !Regressed table correction
  z = -0.803 + 1.312 * x - 0.2118 * x * x + 0.016 * x * x * x
  goto 200
100 x = 1.0 - x;
  !Guard against zero discontinuity
  if (x <= 0) goto 210
  x = -dlog(x)
  z = 0.803 - 1.312 * x + 0.2118 * x * x - 0.016 * x * x * x 
200 x = 2.0 / (9.0 * m)
  x = 1.0 - x + z * dsqrt(x)
210 x = m * x * x * x
  return
end


! End of file chisqa.f90