How to modify this code for calculation seasonal values for each year (not all years)

I have a 1x3 cell contains three 360x3 tables. For every table, I want to calculate the average of rrr24 for each season and write them on the 5th column. For example:

I have a code as you can see below, but it calculates the average of each season for all 30 years (not yearly), and didn't write averages on the 5th column too.

I would like just a new 1x3 cell with tables that have the 5th column as I described above.

meanTableDataArray = cell(length(CELL),4);
for j = 1:length(CELL)
    sampleTableData = CELL{j};
    sampleTableData.month = month(sampleTableData.dates);
    sampleTableData.seasons = floor(sampleTableData.month ./3);
    sampleTableData.seasons(sampleTableData.seasons ==4 ) = 0;
    sampleTableData.seasons = categorical(sampleTableData.seasons, [0 1 2 3], ["Spr", "Sum", "Aut", "Win"]);
    [group, mean_table] = findgroups(sampleTableData(:, 'seasons'));
    mean_table.rrr24 = splitapply(@mean, sampleTableData.rrr24, group);
    for k = 1:4
        meanTableDataArray{j,k} = mean_table(k,:); % extracts as a table
    end
end

Thank you.

[darova]

I like your drawing

Answer 1

[Ameer Hamza]

Try this

data = load('CELL.mat');
CELL = data.CELL;

month_names = month(datetime(1,1,1):calmonths(1):datetime(1,12,1), 'name');
seasons = ["Spr", "Sum", "Aut", "Win"];

for i=1:numel(CELL)
    CELL{i}.grid_name = [];
    CELL{i}.month = month_names(month(CELL{i}.dates))';
    CELL{i}.seasons = seasons(floor((month(CELL{i}.dates)+2)/3))';
    grps = findgroups(findgroups(year(CELL{i}.dates), CELL{i}.seasons));
    avg_value = accumarray(grps, CELL{i}.rrr24, [], @mean);
    CELL{i}.Average = avg_value(grps);
end

[BN]

Dear Ameer Hamza,

Thank you.

Just a little problem. I saw that the code classify January and February (and other months) in the wrong season, for example January in the Spring. I know that this is my problem because in the first code in my question I typed Spring first, so I changed your code to this:

month_names = month(datetime(1,1,1):calmonths(1):datetime(1,12,1), 'name');
seasons = ["Win", "Spr", "Sum", "Aut"];

But it is not accomplished and this time for example code considers March in the winter.

I thought about this code today about 8 hours; thought I could solve this problem if I add an extra row above all rows with NaN value for rrr24 and 12/01/1988 for dates column but even with this method, It is not fixed.

Furthermore, since it is not possible to have a table with the different number of rows for each column I want to ask you if you please guide me how I can store the output of the code in another cell in order to have only one number for each season:

Thank you so much again.

[Ameer Hamza]

Behzad, for first question, can you clarify how do you want to classify the months into 4 seasons. Is this correct:

Jan to March -> Winter

April to June -> Spring

July to September -> Summer

October to December -> Aut

Or is there some other classification for the months in a season.

For second question. As you mentioned that it is not possible to have different number of rows in a table. So do you want to save it as a cell array in a seperate variable?

[BN]

I want this classification:

Winter = December - January - February

Spring = March - April - May

Summer: June- July - August

Autumn = September - October - November

______________________________________

Yes I need to save it in a new 1 x 3 cell with tables as arrays.

Really thank you.

[Ameer Hamza]

Try this

data = load('CELL.mat');
CELL = data.CELL;
month_names = month(datetime(1,1,1):calmonths(1):datetime(1,12,1), 'name');
seasons = ["Win", "Spr", "Sum", "Aut"];
avg_Results = cell(1,3);
for i=1:numel(CELL)
    CELL{i}.grid_name = [];
    CELL{i}.month = month_names(month(CELL{i}.dates))';
    CELL{i}.seasons = seasons(ceil((mod(month(CELL{1}.dates+1),12)+1)/3))';
    grps = findgroups(findgroups(year(CELL{i}.dates), CELL{i}.seasons));
    avg_value = accumarray(grps, CELL{i}.rrr24, [], @mean);
    avg_Results{i} = avg_value(unique(grps, 'stable'));
end

[BN]

I really appreciate you. Thanks!

[Ameer Hamza]

Glad to be of help.