Sums from different groupings

Dear all. I have the table below:

n	x1	x2	x3	x4      x5    x6    ...
1	20	9	166	124    
2	21	8	80	106    
3	23	6	94	105     
4	25	7	22	90
5	23	9	166	82
6	21	8	100	72
7	24	8	375	82
8	22	9	350	88
9	25	7	84	105
10	21	5	22	90
11	22	9	59	124
12	26	7	51	106
13	23	6	51	105
...				
1118				

The x1, x2 and x3 columns are collected data. The x4 is a calculus using the x1 and x2 columns.

I need to sum in 3 to 3, 6 to 6, 9 to 9, 12 to 12, 18 to 18, 24 to 24, 48 to 48... the columns x3 and x4 and rearrange the partial sums in the x5, x6, x8... columns.

Results from x5 and x6 need to be referent to 3 to 3; x7 and x8 to 6 to 6 and so on.

How can I solve this?

Thks!

Answer 1

[Rik]

You can pad a grouping with NaN values and still use a similar idea as your other question:

x=1:17;
N=3;
if mod(numel(x),N)~=0
    x((numel(x)+1):N*ceil(numel(x)/N))=NaN;
end
x
squeeze(sum(reshape(x,3,1,[]),'omitnan'))

[rpid]

Thank you, it helped me a lot! All groupings tested, problem solved.