# How to write a LOOP for this case?

1 view (last 30 days)
BN2 on 28 Oct 2019
Commented: BN2 on 28 Oct 2019
Hey all
I have 34 variable in my workplace like this:
precip_1982_daily
precip_1983_daily
precip_1984_daily
precip_1985_daily
...
precip_2015_daily
I want to use this if code below:
if size(precip_1982_daily,3)==365
month_lengths=[31 28 31 30 31 30 31 31 30 31 30 31];%non-leap year
else
month_lengths=[31 29 31 30 31 30 31 31 30 31 30 31];%leap year
end
precip_1982_monthly=mat2cell(precip_1982_daily,size(precip_1982_daily,1),size(precip_1982_daily,2),month_lengths);%divide into 1 cell per month
precip_1982_monthly=cellfun(@(x) nansum(x,3),precip_1982_monthly,'UniformOutput',false);%find mean for each month
precip_1982_monthly=cell2mat(precip_1982_monthly);%convert back to 3D array

Guillaume on 28 Oct 2019
So again, numbered variables and field names are a bad idea. (see here and here) for the related questions). The whole thing would be simpler if the original cell array was used.
If the original cell array had been kept, the code over the 3 questions could be simplified to:
%using original cell array precips
precip_monthly = cell(size(precips)); %preallocate result cell array
for idx = 1:numel(precips)
if size(precips{idx}, 3) == 365
month_lengths=[31 28 31 30 31 30 31 31 30 31 30 31];%non-leap year
else
month_lengths=[31 29 31 30 31 30 31 31 30 31 30 31];%leap year
end
p_month =mat2cell(precips{idx}, size(precips{idx}, 1), size(precips{idx}, 2), month_lengths);%divide into 1 cell per month
p_month = cellfun(@(x) sum(x, 3, 'omitnan'), p_month, 'UniformOutput', false); %find mean for each month !!THIS CALCULATE THE SUM!!
precip_monthly{idx} = cell2mat(p_month);%convert back to 3D array
end
Note that you have written a comment that says find mean for each month but you use the sum function. So either the code or the comment is wrong.
BN2 on 28 Oct 2019
special thanks to both of you :)

Shubham Gupta on 28 Oct 2019
Edited: Shubham Gupta on 28 Oct 2019
I think this is the continuation of this question.
So, once you create struture as mentioned in the above answer, you can use:
for i = 1:length(precip)
precip_daily = precip.(['y',num2str(i+1981)]); % Access data of specific year
if size(precip_daily,3)==365
month_lengths=[31 28 31 30 31 30 31 31 30 31 30 31];%non-leap year
else
month_lengths=[31 29 31 30 31 30 31 31 30 31 30 31];%leap year
end
precip_monthly=mat2cell(precip_daily,size(precip_daily,1),size(precip_daily,2),month_lengths);%divide into 1 cell per month
precip_monthly=cellfun(@(x) nansum(x,3),precip_monthly,'UniformOutput',false);%find mean for each month
precip_monthly=cell2mat(precip_monthly);%convert back to 3D array
precp_monthly.(['y',num2str(i+1981)]) = precip_monthly; % Create precip_monthly structure
end
BN2 on 28 Oct 2019
Thank you, it's perfect code.