Calculate Eigen Vector and Eigen Value for PCA

asked 2017-01-20 21:45:22 -0600

Shabrina
16 ●1 ●3

I want to get eigenvalue and eigenvector. But it breaks. this is my part of code.

Mat covmat=((matdev*matdev.t())/3); //data float matrix 5*5, this's working. but when I add with PCA function. It's break.
PCA pt_pca(covmat, Mat(), CV_PCA_DATA_AS_ROW, 0);
            Mat pt_eig_vals = pt_pca.eigenvalues;
            for (int i = 0; i < 4; ++i)
            cout << pt_eig_vals.at<float>(i, 0) <<endl;

This is the error. OpenCV error: Assertion Failed <data.channerls<> =="1"> in unknown function, file ......\src\opencv\modules\core\src\matmul.cpp

Please somebody help me :(. Thanks

edit retag flag offensive close merge delete

Comments

what is matdev ? (please show, how you construct that)

can you try again to retrieve the correct error msg ? data.channels()==1 would make sense, data.channels<>==1 is a syntax error.

berak ( 2017-01-21 01:02:51 -0600 )edit

matdev is matrix data from txtfile like this.

%YAML:1.0
Point: !!opencv-matrix
   rows: 5
   cols: 5
   dt: "2f"
   data: [ 0., 0., 1., 0., -1., 0., 0., 0., 0., 0., 0., 0., -1., 0., -1.,
       0., 0., 0., -1., 0., 0., 0., 0., 0., 1., 0., -1., 0., 0., 0., 1.,
       0., 1., 0., -1., 0., 0., 0., -1., 0., 0., 0., 0., 0., 0., 0., 0.,
       0., 0., 0. ]

belan ( 2017-01-21 01:14:45 -0600 )edit

maybe it helps, if you could also explain, how you got this, and what is inside your data ?

berak ( 2017-01-21 01:41:45 -0600 )edit

The data is vector x y -> 25 data. and I change/ reshape it to nxn = 5x5 (this is int). after that I convert to 32F. and that the result. any i idea for change it to 1 channel matrix with row n column same . thank you

belan ( 2017-01-21 08:02:25 -0600 )edit

add a comment

Comments

So, how to make it, to be single channel? I'm a beginner in opencv.

belan ( 2017-01-21 01:30:11 -0600 )edit

you could either split it into 2 planes, and handle those seperately:

Mat planes[2];
split(matdev, planes);
PCA pca1(planes[0],...);
PCA pca2(planes[1],...);

or try to reshape it to a 1channel, [10x5] mat:

Mat matdev1 = matdev.reshape(1,5);

but i have no idea, if any of this makes sense with your data. maybe you have to go back, andfind out, why matdev looks like it is, and if this is really the desired format.

berak ( 2017-01-21 01:40:22 -0600 )edit

Ok, thank you, I will try it.

belan ( 2017-01-21 06:14:26 -0600 )edit

add a comment

Calculate Eigen Vector and Eigen Value for PCA

Comments

1 answer

Comments

Links

Question Tools

Stats

Related questions

Calculate Eigen Vector and Eigen Value for PCA edit

Comments

1 answer

Comments

Links

Question Tools

Stats

Related questions

Calculate Eigen Vector and Eigen Value for PCA