Which should I use for this integration.

Hello everyone

I have to make a function to calculate significance of given arrays (lets say FC = C12 which is 47X1 double)

The fucnction should look like

function alpha = significance (FC)
% Variable FC is consistant values ​​coloum. For example FC = C12 will give the significane of C12 coloum.

n = length (FC); % Calculates the Number of consistent respondents (Nt)
x = sum (FC == mode (FC)); % Calculates the Number of respondents answered mode
p = 1/17; % Probability of selecting mode (Pi)
mu = n * p; 
mu1 = n * (1-p);
sigma = sqrt (n. * p. * (1 - p)); % Standerd deviation

if n <= 20
    i = 0;
    syms i
    alpha = symsum ((factorial (n) / (factorial (i) * factorial (ni)) * (1/17). ^ (i) * (1-1 / 17). ^ (ni)), i, 0 , x);
elseif n> 20
    syms y
    f = (- (y-mu). ^ 2) / (2 * sigma. ^ 2); % 5.05 is 2 * sigma ^ 2, & 2.7 = mu
    fun = (1 / (sigma * 2.5)) * (exp (f)); % 0.2516 = 1 / (sigma * 2.5), e = 2.71
    a = -inf;
    b = x; % x; % number of respondent selected mode
    alpha = int (fun, a, b);
  %%% ^ Which Integration function should I use here?
end

The problem here is, when I am running the script in commond window seperatly with values ​​given as comment, it give me the currect value.

>> syms y
    f = (- (y-2.7). ^ 2) / (5.05); % 5.05 is 2 * sigma ^ 2, & 2.7 = mu
    fun = (0.2516) * (exp (f)); % 0.2516 = 1 / (sigma * 2.5), e = 2.71
    a = -inf;
    b = 19; % x; % number of respondent selected mode
    alpha = int (fun, a, b);
>> alpha
 
alpha =
 
(629 * 2020 ^ (1/2) * pi ^ (1/2) * (erf ((163 * 2020 ^ (1/2)) / 1010) + 1)) / 100000
 
>> ans = (629 * 2020 ^ (1/2) * pi ^ (1/2) * (erf ((163 * 2020 ^ (1/2)) / 1010) + 1)) / 100000

ans =

    1.0021

But when I run it in function above, I am getting my answer '0',

For FC = C12,

n = 47

x = 19

I want to know if I am using currect integration function? If not then which one should I use.

Please currect me.

[Ameer Hamza]

Can you show us the lines fo code you are using to integrate significance.

[Ashwani Kumar MAlviya]

Thank Ameer for your response.

here is the code I want to integrate alfa = int(...)

elseif n> 20
    syms y
    f = (- (y-mu). ^ 2) / (2 * sigma. ^ 2); % 5.05 is 2 * sigma ^ 2, & 2.7 = mu
    fun = (1 / (sigma * 2.5)) * (exp (f)); % 0.2516 = 1 / (sigma * 2.5), e = 2.71
    a = -inf;
    b = x; % x; % number of respondent selected mode
    alpha = int (fun, a, b);
  %%%       ^^ Which Integration function should I use here?
end

[Navya Seelam]

Hi,

The integral function you are using is correct. I recommend to recheck your logic.

[Ashwani Kumar MAlviya]

Thanks Navya.