how to put an if statement within for loop

I compile climate data over many years by month within a for loop, and concatenate these to create a 3D array (of say January's over many years), and I want to take the element-wise mean to get the average over those years:

monthlydata = cell(1, numel(years));  %to hold the monthly data of each year. Preallocated.
years = 2003:2014; 
for month = 1:12
   for y = 1:numel(years)
      monthlydata{y} = load(sprintf('sst_%d_%d.asc', years(y), month));  
   end 

   monthlydatac = cat(3, monthlydata{:});  %concatenate all cells along 3rd dimension
   monthlymean = mean(monthlydatac, 3);    %get mean along 3 dimensions
   dlmwrite(sprintf('msst_%d.csv', month), monthlymean);  %save

end

The problem is there is missing data for some months taking the form of numbers 250, 251, 252, 253, 254, 255. In some years observations occur in elements of the 3D array but in other years data is missing ( coded as 250:255) for that element. If this is the case I want to take the element-wise mean excluding these 'missing data' values. Would an 'if' statement work? Could anyone advise me on how to do this?

I am new to Matlab and programming and would be very greatly for your help.

[Jan]

I do not understand "missing data in some elements of some months in the form of 255, 254, 253, 252, 251 or 250". What is missing? The file or the data? What does "In some years real observations occur where usually it they are missing" mean? What does it mean, that you want to "specify" this? What is "take whatever observation is real" exactly?

Please edit the question and define exactly, what you want to do with which values. I'm sure it will work e.g. with an if command. But for an explicit suggestion, it is required to know, what you want.

[Roisin Loughnane]

The data is missing for some years and coded as one of those numbers 250:255, but for some of those same elements some years may have a value I would like to keep.

I hope the edit helps. Thank you

[Jan]

No, this does not help. What is "coded as one of those numbers 250:255"? How could we suggest an explicit solution based on the detail, that "some" elements of "some" years "may" have a value, that you "would like" to keep?

An exact, unequivocal, clear and unique definition is required, to implement this. We cannot do this for you. I assume the actual implementation will be very easy then.

[Roisin Loughnane]

NaN = a number between 250 and 255 in my dataset.

For some arrays in the 3D array there are NaN values in some elements, although in corresponding elements in different arrays there is data I am interested in.

The mean is being calculated including a missing data value (here a number between 250 and 255) which I do not want.

I want to exclude any value between 250:255 before getting the mean of the rest of the corresponding elements of the 3D array.

Answer 1

[Rik]

Time for a pure guess: you want to ignore values in the 250-255 range and calculate the mean over the remaining data?

monthlydatac(monthlydatac>=250 & monthlydatac<=255)=NaN;
monthlymean = mean(monthlydatac, 3, 'omitnan');

[Roisin Loughnane]

Thank you very much.

Sorry for the confusion.

[Jan]

@Rik: +1, a good guess.