how to do mean of a column with a loop and excluding nans?

so i have this large file that is contains nans randomly, its a 1140 X 1 matrix i want to do a for loop that excludes all the nans and do the mean of that can someone explain me how to do this?

[tonyrodriguez]

i cant use nanmean for this

Answer 1

[Birdman]

You do not need a for loop. Matlab's mean function has an option of omitnan. You can use it as follows:

res=mean(a,'omitnan')

[tonyrodriguez]

i know but i need to use a for loop to do this... i cannot just use the mean function directly.. or the nanmean i need to use a for loop to exclude all the nans first .. then i can do the mean..

[Birdman]

Then:

res=0;cnt=0;
for i=1:numel(a)
   if ~isnan(a(i))
      res=res+a(i);
      cnt=cnt+1;
   end
end
ave=res/cnt

[tonyrodriguez]

that works thank you! .. but can you explain me what the res do? can you do without it? and just count? since in my class we have only used count... and perhaps do at the end mean(isnan(a(i)) ?

[Birdman]

res is used to sum up the variables that are not NaN, that is it. You need a variable to sum up the variables that are not NaN in a for loop, therefore you need it and cnt simply counts how many times that if statement is entered.