Simpler code for aggregating data by sum

2 次查看（过去 30 天）
Clarisha Nijman
Clarisha Nijman 2019-1-10
评论： Clarisha Nijman 2019-1-12
Hello,
given matrix A: A=[4 0.5;4 0.25;1 0.125; 2 0.2;4 0.6;3 0.2; 1 2/3; 2 1; 2 1/16;4 0.5]
asked: Aggregate data in first column by sum
desired result:
1.0000 0.7917
2.0000 1.2625
3.0000 0.2000
4.0000 1.8500
I composed a code, but it work half, it can not make the data unique. So it fails at : G=unique(List)
Can somebody tell me why it fails at that point? And is there a more straight forward code for this whole proces? Maybe inbuilt functions? Can I get some feedback pls?
This is my code:
List=[];
% Make list unique
for i=1:size(A,1)
F=A(A(:,1)==A(i,1),:); %subsetting all rows
ElementList=[A(i,1) sum(F(:,2))]; %adding the sum of the subset to a row
List=[List; ElementList] % saving row plus its aggregated sum to a list
end
G=unique(List) %making the list unique
kind regards,
  3 个评论
显示 1更早的评论
Akira Agata
Akira Agata 2019-1-10
Or simply:
A = [4 0.5;4 0.25;1 0.125; 2 0.2;4 0.6;3 0.2; 1 2/3; 2 1; 2 1/16;4 0.5];
B = splitapply(@sum,A(:,2),A(:,1));
The result is:
>>[unique(A(:,1)), B]
ans =
1.0000 0.7917
2.0000 1.2625
3.0000 0.2000
4.0000 1.8500
Clarisha Nijman
Clarisha Nijman 2019-1-11
Thank you dear friends!!! Suggestion woks perfect!

请先登录，再进行评论。

请先登录，再回答此问题。

采纳的回答

OCDER
OCDER 2019-1-10
A=[4 0.5;4 0.25;1 0.125; 2 0.2;4 0.6;3 0.2; 1 2/3; 2 1; 2 1/16;4 0.5]
Results = [unique(A(:, 1)), accumarray(A(:, 1), A(:, 2))];
Results =
1.0000 0.7917
2.0000 1.2625
3.0000 0.2000
4.0000 1.8500
  4 个评论
显示 2更早的评论
OCDER
OCDER 2019-1-11
Check your results, as you have [3 4 2] unique combo.
A=[2.0000 3.0000 4.0000 0.01;
2.0000 3.0000 4.0000 0.05;
3.0000 2.0000 4.0000 0.01;
3.0000 4.0000 2.0000 0.02;
1.0000 2.0000 3.0000 0.03;
2.0000 3.0000 4.0000 0.00;
1.0000 2.0000 3.0000 0.02;
2.0000 3.0000 4.0000 0.01;
2.0000 3.0000 4.0000 0.01;
3.0000 2.0000 4.0000 0.07];
[UnqVal, ~, UnqGrpNum] = unique(A(:, 1:3), 'rows');
E = [UnqVal, accumarray(UnqGrpNum, A(:, end))];
E =
1.0000 2.0000 3.0000 0.0500
2.0000 3.0000 4.0000 0.0800
3.0000 2.0000 4.0000 0.0800
3.0000 4.0000 2.0000 0.0200
Clarisha Nijman
Clarisha Nijman 2019-1-12
Ok I see, you have to assign the unique groups a number, such that you can refer to them in the accumarray function. If would never come up with such a good idea. I worked with for loops and the code runs very very long. I indeed missed the unique combo[3 4 2] in my results.
Thanks a lot!!!

请先登录，再进行评论。

更多回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Characters and Strings 的更多信息

标签

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by