How to find the index of splitapply

Question

Yaser Khojah 2019-4-29

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/459199-how-to-find-the-index-of-splitapply

评论： Yaser Khojah 2019-4-30

I have this code where I split the data into intervals and I look to find the maximum value for each interval as below. I need to find the indexes of the maximum values (Result) as well to collect more information from the original matrix (data). For simplify I created the following example.

x          = round(Data(:,18),0);
y          = Data(:,17);
[uv,~,idx] = unique(x);
[Result]   = splitapply(@max, y, idx);

How can I find the indexes correspond to the Result to get more information from the original data.

Thank you so much for your help.

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Matt J 2019-4-29

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/459199-how-to-find-the-index-of-splitapply#answer_372823

编辑：Matt J 2019-4-29

在 MATLAB Online 中打开

The indices within each group or the indices over all of y? If the former,

indices   = splitapply(@maxIndex, y, idx);
function out=maxIndex(z)
  [~,out]=max(z);
end

7 个评论
显示 5更早的评论隐藏 5更早的评论

Yaser Khojah 2019-4-30

编辑：Yaser Khojah 2019-4-30

在 MATLAB Online 中打开

Dear Matt,

I'm not really sure why I get these values for the indices? they are all ones as below although uv and Result are different numbers as below:

M = [uv Result indices]
-52.6375201767288	1
8.16048927046429	1
32.8959501232429	1
40.6235344557864	1
34.9060598363029	1
40.1999523692772	1
40.8218054895630	1
40.6603648293852	1
44.9101394298133	1
46.3799406552174	1
45.5282239893887	1
44.5752434568953	1
43.9269340744265	1
47.2730139688185	1
47.5348218152510	1
47.2414263217449	1
46.1472661513097	1
43.8672894544233	1
43.9554365031925	1
43.3286228656092	1
40.6979098991405	1
39.9119400060329	1
40.4076381192138	1
40.6247235584278	1
28.6474646180134	1
24.1315924439042	1
25.8657575418280	1
6.87298386234900	1
8.86434344246124	1
14.9836953775187	1
11.0986590169539	1
-13.2860150017084	1
5.28578052709843	1
8.46536624907806	1
2.08700412241725	1
2.33072961721340	1
5.64500984492813	1
-7.82948439124770	1
-20.4453841342906	1

If you try it on this example you will see the indices are not correct:

XX = [1   2  3  4  5  1  3]'; 
YY = [10 10 10 20 20 30 20]';
[uv,~,idx] = unique(XX);
[Result]   = splitapply(@max, YY, idx);
figure
scatter(uv,Result)
line(uv,Result)
indices   = splitapply(@maxIndex, YY, idx);
function out=maxIndex(z)
  [~,out]=max(z);
end
m = [uv, Result, indices]
     1    30     2
     2    10     1
     3    20     2
     4    20     1
     5    20     1
 
%The correct one should be     
m = 
1	30	6
2	10	2
3	20	8
4	20	4
5	20	5
%any idea how to fix this?

Matt J 2019-4-30

编辑：Matt J 2019-4-30

在 MATLAB Online 中打开

That is why I asked you at the beginning if you meant, "The indices within each group or the indices over all of y?" If you meant the former, then your first set of results is correct. For example, if I extract all the YY for which idx==1, I obtain,

group=[10,30]

and clearly the maximum over the group occurs at the second element, so indices=2 is the correct output.

However, since It is now clear that you meant "the indices over all of y", here is the solution for that,

XX = [1   2  3  4  5  1  3].'; 
YY = [10 10 10 20 20 30 20].';
[uv,~,idx] = unique(XX,'stable');
II=(1:numel(XX)).';
S  = splitapply(@(xy)maxIndex(xy), [II,YY], idx);
[Result,indices]=deal(S(:,1),S(:,2))
m = [uv, Result, indices]
function out=maxIndex(xy)
  [ymax,loc]=max(xy(:,2),[],1);
  out=[ymax,xy(loc,1)];
end