Why I get ans = logical 1

for i=1:100 
    x(i)=normrnd(0,1); 
end; 
A=x(:);
for i=1:100 
    if (A(i)<0.9557 & A(i) > 0) 
        Aq(i)=0.497; 
    elseif (A(i)>0.9957) 
        Aq(i)=1.493;
    elseif (A(i)<-0.9957) 
        Aq(i)=-1.493; 
    else 
        (A(i)>-0.9957 & A(i) <0) 
        Aq(i)=-0.497;
    end; 
end;
sum=0; 
for i=1:100
    sum = (A(i)-Aq(i))^2+sum; 
end; 
Avg = sum/100

[Geoff Hayes]

Mohamed - look closely at your else

    else 
        (A(i)>-0.9957 & A(i) <0) 
        Aq(i)=-0.497;
    end; 

Should this be an elseif instead where (A(i)>-0.9957 & A(i) <0) is the condition?

[Mohamed Mahir]

Thank you very much it fixed it

Answer 1

[Steven Lord]

Another approach uses discretize.

>> A = randn(10, 1);
>> edges = [-Inf -0.9957 0 0.9957 Inf];
>> values = [-1.493, -0.497, 0.497, 1.493];
>> Aq = discretize(A, edges, values);
>> results = table(A, Aq)

If an element of A falls between (for example) edges(2) and edges(3) the corresponding element in Aq will be values(2).

I put the results in a table array so you can easily check that each element of Aq matches what it should be for the corresponding element of A.

[Mohamed Mahir]

Thank you that really helped me

Answer 2

[David Hill]

Much easier way:

A=normrnd(0,1,[100,1]);
[~,~,a]=histcounts(A,[-10,-.9957,0,.9957,10]);
b=[-1.493;-0.497;0.497;1.493];
Aq=b(a);
s=sum((A-Aq).^2);
Avg = s/100;

[Mohamed Mahir]

Thank you