Algorithm and pseudo code for the calculation

Introduction

Objectives

Methodology

Algorithm and pseudo code for the calculation

For n = 1 to 200

Do 1:1000000

Rank1 , Rank2  Rankfsubfunction(List(n))

W  calTvale(Rank1 , Rank2)

α = .01 , Value2 = SortedRanList (99% of million )

α = .025 , Value3 = SortedRanList (98% of million )

Implementation design and formula

General formula for two tailed t vale

Tc α = (1 – 100* α)^th element in the sorted list

i<= n/2 i=>n/2

Solution and explanation

ALLOCATE(x(n))

ALLOCATE(y(n))

ALLOCATE(Ta((Iteration+1)/2))

Million times iteration so that we can get million array of size n

x(i)=z(i)

y(i)= z(n+i)

call calTvale(R , S , n, t)

wilx = t

END IF

end do

deallocate (R)

deallocate (S)

stop

end program using_uni1

The Implementation Results

Table of critical values for Maximum T-Test and Wilcoxon

--------------------------------------------------------------------------------------

df al=.001 al=.01 al=.025 al=.050

8 14.0267 4.00000 3.25553 2.53287

10 9.59463 3.86513 3.00000 2.27863

20 9.00000 3.00000 2.52236 2.00000

22 8.12252 2.99756 2.47980 2.00000

32 6.54099 2.76151 2.35685 2.00000

48 6.00000 2.65512 2.29808 2.00000

70 6.43069 2.59588 2.26301 2.00000

78 5.28118 2.59392 2.26094 2.00000

120 5.15038 2.53686 2.00000 2.00000

122 5.00000 2.53461 2.22462 2.00000

214 5.00000 2.53031 2.22280 2.00000

220 5.00000 2.52732 2.22089 2.00000

Limitation and further improvement

We used two tailed for equal sample of data for each list
Due to time complexity we are not able to go for very big number of samples though there would not be a significant deviation

Better sorting algorithm like external sorting to handle big data
Faster process like super computer could be used