code for computing the formular

I have the set of numbers

(Si)  = (0.25541156, 4.446482251, 0.389762062, 4.211214131)
(sj) = ( 0.25541156, 4.446482251, 0.389762062, 4.211214131)
 pvalue =     ( #( si greater than sj )   +   0.5 (si  = sj) )/ i

when i = 1

I am trying to compute this steps

pvalue = 1  when i = 1  pvalue is always 1
si = sj = 1

when  i = 2
J = 1         if  s2 > sj1 = 1        pvalue = ( 1 + 0.5(1)) / 2 = 0.75
J = 2             s2 = sj2 = 1 

for i = 3
J = 1         s3 > sj1 = 1
J = 2         s3 > sJ2 = 0            pvalue = (1 + 0.5 (1))/3  =   0.5
J= 3          s3 = sj3 = 1

for i = 4
J = 1          s4 > sj1 = 1
J = 2          S4 > Sj2 = 0           pvalue = 2 + 0.5(1)/4  = 0.625
J = 3          s4 > sj3 = 1
J = 4          s4 = sj4 = 1

I will appreciate it if I get a code to compute the various pvalue

jonathan

Answer 1

[dpb]

fnP=@(a,i) (sum(a(i)>a(1:i))+0.5*sum(a(i)==a(1:i)))/i;

>> for i=2:numel(Si),fnP(Si,i),end
ans =
          0.75
ans =
          0.50
ans =
          0.63
>> 

To wrap the i==1 special case got more than I could get into the anonymous function in the time I had to play...for general use write a little function--

function P=fnP(A,n)
% Return some undefined P-value estimator from vector A, number elements, n

  if n==1
    P==1;
  else
   P=(sum(A(n)>A(1:n)) + 0.5*sum(A(n)==A(1:n)))/n;
  end
end