Issues with the for loop

tt=0;
for i = 1:size(ids_2006,2) * size od ids_2006 is 212*119*177
    for j = 1:size(ids_2006,3)
            tt=tt+1;
            i,j,tt
            % size of mean_wint_rho is 119*177
            % size of stdv_daily_2006 is 119*177
        if (squeeze(ids_2006(tt,:,:)) < (mean_wint_rho(:,:) - 3*stdv_daily_2006(:,:)))
            daily_2006 = squeeze(ids_2006(:,i,j));
            new_daily_2006(tt,:,:) = daily_2006; 
            %size of new_daily_2006 should be 212*119*177
        else
            new_daily_2006(tt,:,:) = NaN;
        end
    end
end

I want to remove the bias from the daily dataset by removing the points that are three standard deviation from the mean.

Answer 1

[darova]

This should work

new_daily_2006 = ids_2006*0;
for i = 1:size(ids_2006,1)  %* size od ids_2006 is 212*119*177 
%     i,j,tt
    % size of mean_wint_rho is 119*177
    % size of stdv_daily_2006 is 119*177
    if (squeeze(ids_2006(i,:,:)) < (mean_wint_rho(:,:) - 3*stdv_daily_2006(:,:)))
        new_daily_2006(i,:,:) = squeeze(ids_2006(i,:,:));
        %size of new_daily_2006 should be 212*119*177
    else
        new_daily_2006(i,:,:) = NaN;
    end
end

The question is: when it will be work?

if (squeeze(ids_2006(i,:,:)) < (mean_wint_rho(:,:) - 3*stdv_daily_2006(:,:)))

Something like this

if ([1 0])
    disp('hi')
end

Or maybe this

if any([1 0])
    disp('hi')
end

[Sophia]

I understood your point regarding plotting. But other than that i tried the following and its is not correct, or what i was trying for.

new_daily_2006 = [];
 
for i = 1:size(ids_2006,1)
    if (abs(squeeze(ids_2006(i,:,:))) < 3 * stdv_daily_2006(:,:))
        new_daily_2006(i,:,:) = squeeze(ids_2006(i,:,:));
    else
        new_daily_2006(i,:,:) = NaN;
    end              
end

new_daily_2006 is 212*1, i was expecting it to be 212*119*177

[Sophia]

I have finally figured the answer-

new_daily_2006 = [];
 
for t = 1:size(ids_2006,1)
   for i = 1:size(ids_2006,2)
     for j = 1:size(ids_2006,3)       
         temp = squeeze(ids_2006(t,i,j));
            if (abs(temp) < stdv_daily_2006(i,j))
                new_daily_2006(t,i,j) = temp;
            else
                new_daily_2006(t,i,j) = NaN;
            end  
        end
    end
end

[darova]

Try also without loops

m = size(ids_2006);
temp = reshape(stdv_daily_2006,[1 size(stdv_dayly_2006)]); % make 1 x 119 x 177 
cond = abs(ids_2006) < repmat(temp,[m 1 1])              % make 212 x 119 x 177
new_daily_2006(cond) = ids_2006;
new_daily_2006(~cond) = NaN;