average each block of class in a matrix
2 次查看(过去 30 天)
显示 更早的评论
Hi, I have a large matrix where the rows are labelled according to different classes. I would like to obtain an average row of each class. How can I do it?
Thank you!
0 个评论
采纳的回答
Guillaume
2020-3-16
编辑:Guillaume
2020-3-16
It's unclear how your input data is actually stored in matlab so please clarify if the below doesn't apply.
I'm assuming that you have a two column matrix with one column storing the the class and the other the column storing the values you want to average by class. In this case, this is trivially achieved with groupsummary:
classcolumn = 1; %change as appropriate
valuecolumn = 2; %can be more than one column. Each column will be average by class
classmean = groupsummary(yourmatrix, classcolumn, 'mean', valuecolumn);
classcolumn = 1;
valuecolumn = 2;
group = findgroups(yourmatrix(:, classcolumn));
classmean = splitapply(@mean, yourmatrix(:, valuecolumn), group);
%or
classmean = accumarray(group, yourmatrix(:, valuecolumn), [], @mean); %this one can only work on one column though
0 个评论
更多回答(1 个)
Sriram Tadavarty
2020-3-16
Hi Alessio,
The mean function will help in this case. Here is the documentation page of mean function having different exampleshttps://www.mathworks.com/help/matlab/ref/mean.html
% Consider a random matrix
a = randn(1000,10000); % Implies 1000 rows and each row containing 10000 columns
% Calculate the mean of each column
meCol = mean(a);
% Calculate the mean of each row
meRow = mean(a,2); % This is your intended case
Hope this helps.
Regards,
Sriram
2 个评论
Sriram Tadavarty
2020-3-16
Hi Alessio,
In this case, you can run a fop loop on number of rows and take mean for each row.
% Rows variable is a cell array with different number of elements in it. For example
Rows = {[1 4 5],[1 9 10 100]};
for i = 1:numel(Rows)
avg(i) = mean(Rows{i});
end
% avg returns [3.3333 30.0000] each element corresponding to each row
Hope this helps
Sriram
另请参阅
类别
在 Help Center 和 File Exchange 中查找有关 Data Preprocessing 的更多信息
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!