simplify my script?

hello, I need some suggestion

I know my code isn't perfect, but this is enough for me. just wondering if my code could be more compact or simpler. the point is the same only different in the grouping per 10 days. I only want to run my code once. here is my code.

for n=1:12
    % Load data
    ncfile = 'Temperature.nc';
    months = {'January', 'February', 'March', 'April', 'May', 'June', ...
        'July', 'August', 'September', 'October', 'November', 'December'};
    
    % to read a variable 'var' exisiting in nc file
    lon = ncread(ncfile,'longitude');
    lat = ncread(ncfile,'latitude');
    tempI = ncread(ncfile,'t2m');
    
    % grup date
    startDate = datetime(2021,1,1,0,0,0);
    endDate = datetime(2021,12,31,23,0,0);
    dates = startDate:hours(1):endDate;
    
    %% pick each 10 days data
    Onemonth = dates.Month==n ;
    t = dates(Onemonth) ;
    Ddays = t.Day<=10 ; %date 1-10
    % Ddays = t.Day >= 11 & t.Day <=20 ; % date 11-20
    % Ddays = t.Day >= 21 & t.Day <=31 ; % date 21-31
    %%
    % Extract data per month
    t10days = tempI(:,:,Ddays);
    
    % rotate data
    tempII = flipud(rot90(t10days));
    
    % Average data
    temp = mean((tempII-273.15),3);
    
    % Plot
    
    figure
    imagescn(lon,lat,temp);
    colormap jet
    axis image
end
for n=1:12
    % Load data
    ncfile = 'Temperature2m.nc';
    months = {'January', 'February', 'March', 'April', 'May', 'June', ...
        'July', 'August', 'September', 'October', 'November', 'December'};
    
    % to read a variable 'var' exisiting in nc file
    lon = ncread(ncfile,'longitude');
    lat = ncread(ncfile,'latitude');
    tempI = ncread(ncfile,'t2m');
    
    %% grup date
    startDate = datetime(2021,1,1,0,0,0);
    endDate = datetime(2021,12,31,23,0,0);
    dates = startDate:hours(1):endDate;
    
    %% pick each 10 days data
    Onemonth = dates.Month==n ;
    t = dates(Onemonth) ;
    % Ddays = t.Day<=10 ; %date 1-10
    Ddays = t.Day >= 11 & t.Day <=20 ; % date 11-20
    % Ddays = t.Day >= 21 & t.Day <=31 ; % date 21-31
    %%
    % Extract data per month
    t10days = tempI(:,:,Ddays);
    
    % rotate data
    tempII = flipud(rot90(t10days));
    
    % Average data
    temp = mean((tempII-273.15),3);
    
    %% Plot
    
    figure
    imagescn(lon,lat,temp);
    colormap jet
    axis image
    
end
for n=1:12
    %% Load data
    ncfile = 'Temperature2m.nc';
    months = {'January', 'February', 'March', 'April', 'May', 'June', ...
        'July', 'August', 'September', 'October', 'November', 'December'};
    
    % to read a variable 'var' exisiting in nc file
    lon = ncread(ncfile,'longitude');
    lat = ncread(ncfile,'latitude');
    tempI = ncread(ncfile,'t2m');
    
    %% grup date
    startDate = datetime(2021,1,1,0,0,0);
    endDate = datetime(2021,12,31,23,0,0);
    dates = startDate:hours(1):endDate;
    
    %% pick each 10 days data
    Onemonth = dates.Month==n ;
    t = dates(Onemonth) ;
    % Ddays = t.Day<=10 ; %date 1-10
    % Ddays = t.Day >= 11 & t.Day <=20 ; % date 11-20
    Ddays = t.Day >= 21 & t.Day <=31 ; % date 21-31
    %%
    % Extract data per month
    t10days = tempI(:,:,Ddays);
    
    % rotate data
    tempII = flipud(rot90(t10days));
    
    % Average data
    temp = mean((tempII-273.15),3);
    
    %% Plot
    
    figure
    imagescn(lon,lat,temp);
    colormap jet
    axis image
    
end

[Dyuman Joshi]

The 3 loops are identical except for the data file, even then 2nd and 3rd loop loads the same file.

What is your changing variable in all the codes? The file?

You can make your code a function file and call accordingly.

[Den]

yes, all use the same file. therefore I want to make it more compact.

the only difference is the date. first code 1-10, second code 11-20, and 21-31.

I'm new to matlab. just in my understanding, function is used for repetitive for different files. I only use this code once. I just want to know if there is a way to simplify my code

[Dyuman Joshi]

See Jan's answer.

Answer 1

[Jan]

Move all repeated work out of the loops. In your case:

    % Load data
    ncfile = 'Temperature.nc';
    months = {'January', 'February', 'March', 'April', 'May', 'June', ...
        'July', 'August', 'September', 'October', 'November', 'December'};
    
    % to read a variable 'var' exisiting in nc file
    lon = ncread(ncfile,'longitude');
    lat = ncread(ncfile,'latitude');
    tempI = ncread(ncfile,'t2m');
    
    % grup date
    startDate = datetime(2021,1,1,0,0,0);
    endDate = datetime(2021,12,31,23,0,0);
    dates = startDate:hours(1):endDate;

Do this once before all loops, because it does not depend on the loop counter n.

[Johan]

Moreover, the only thing that changes in your loop is the Ddays line, you could define your range of days before the loop and loop for all your wanted ranges:

%Create day_range arrays, first column is low value second is high; each
%row is a new range of selected days
Day_range = [1,10; 11, 20; 21, 31]

Day_range = 3×2

1 10 11 20 21 31

 for i_day = 1:size(Day_range,1)
     for n = 1:12
 %% pick each 10 days data
      Onemonth = dates.Month==n ;
      t = dates(Onemonth) ;
      Ddays = t.Day >= Day_range(i_day,1) & t.Day <= Day_range(i_day,2) ; 
  
      % Extract data per month
      t10days = tempI(:,:,Ddays);
      
      % rotate data
      tempII = flipud(rot90(t10days));
      
      % Average data
      temp = mean((tempII-273.15),3);
      
      %% Plot
      figure(i_day) %new figure for each range
      imagescn(lon,lat,temp);
      colormap jet
      axis image
     end
 end

[Den]

Yes thank you. this is the answer i want.

although I do not understand the logic in this line.

Ddays = t.Day >= Day_range(i_day,1) & t.Day <= Day_range(i_day,2) ; 

but I have matched and the answer is the same.

and if you have time to answer, why use this

 for i_day = 1:size(Day_range,1)

instead of like this

 for i_day = 1:3

What is the difference?

[Jan]

Ddays = (t.Day >= Day_range(i_day,1) & t.Day <= Day_range(i_day,2));

This replies a logical vector, which is TRUE, if the element of t.Day is inside the specified range.

for i_day = 1:size(Day_range,1)    

Of course, if the corresponding size equal 3, you can write 3 also. But this approach is more general and determines the size dynamically. It is a good programming practice to write flexible code without assumptions about the inputs.

Replace

flipud(rot90(t10days));

by the more direct:

permute(t10days, [2, 1, 3]);

Or even better:

t10days = tempI(:,:,Ddays);
tempII  = flipud(rot90(t10days));
temp    = mean((tempII-273.15),3);

by the cheaper and simpler:

temp = mean(tempI(:, :, Ddays), 3).' - 273.15;