Info
This question is closed. Reopen it to edit or answer.
Incorporating Procrustes to PCA
1 view (last 30 days)
Show older comments
Hi:
So I have a PCA matlab GUI. It can perform Jackknifing for displaying a training and prediction cloud plot of the data. The plot doesn't look good because it doesnt perform the rotation well. I want to add procrustes code to it to make it work but i'm finding it difficult in getting around it. The procrustes code is
[d, Z, transform] = procrustes(tdata,tdata)
Below is the code i want to add procrustes to:
colcode_temp = colcode;
if jackState == 2 %training cloud mode
for i = 1:size(tdata,1)
for j = 1:size(jknife,1)
plotMe(size(Xmean,1)+size(jknife,1)*(i-1)+j,1) = jknife(j,i,1);
plotMe(size(Xmean,1)+size(jknife,1)*(i-1)+j,2) = jknife(j,i,2);
colcode_temp(size(Xmean,1)+size(jknife,1)*(i-1)+j,:)=0.7; %turn all cloud points gray
end
end
3 Comments
Walter Roberson
on 26 Sep 2017
If you
[d, Z] = procrustes(tdata,tdata)
you would get back a linear transform -- that is, Z is always going to equal tdata in that case.
You should be using procrustes with two different sets of data.
Answers (0)
This question is closed.
See Also
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!