"Subscript indices must either be real positive integers or logicals" Error... How to solve this!

let ECd

ECd =

   15.7000
   33.9000
   10.2000
    8.4000
   40.7000
   37.4000
   37.2000
   37.8000
   33.7000
   31.6000
   26.7000
   22.9000
   19.6000
    6.1000
    1.9000
    1.5000
    1.4000
    1.2500
    1.8000
    2.3000

N = size(ECd,1);

for i = N:-1:1;

Min2(i-2)= min (ECd(i:N));          
Max2(i-2)= max (ECd(i:N));          
Mean2(i-2)= mean (ECd(i:N));       
Median2(i-2) = median(ECd(i:N));
Std2(i-2)= std (ECd(i:N));          
Q52(i-2)= prctile(ECd(i:N),[5]);
Q752(i-2) = prctile(ECd (i:N),[75]);
end

Its say im having this problem as below

??? Subscript indices must either be real positive integers or logicals.

Error in ==> VarCalculation_ASMS at 93
Min2(i-2)= min (ECd(i:N));          % Min value of ECd

I surf in the community, one of the answer is which size, which cell

I'm not sure what he meant! thanks

Answer 1

[Tom]

the smallest value of i is 1, but you are assigning to an array by using i-2. So for the 2nd to last iteration of the loop you are trying to store in Min2(0). This doesn't exist in Matlab- the array is arranged from 1 to however many you need. This is the 'real or positive integers' part of the error.

[Tom]

the which size/ which cell bit from the other question refers to the fact that you can call a variable the same name as a function. So for example, there is a function called max, which unsurprisingly is used to find the maximum value in a group of numbers. But, if you call a variable 'max', then you can no longer use the max function.

Type 'which max' into the command window, then make a variable called 'max' (e.g. max=1), and type 'which max' again, and see the difference in what comes up.