# A question for principal components analysis (pca)

2 Ansichten (letzte 30 Tage)
Daniel Chou am 15 Feb. 2021
Beantwortet: Gaurav Garg am 22 Feb. 2021
Dear All,
I have a question about principal components analysis.
The scenario is,
I have a ellipsoid with semi major axis (length=1) and two equal length (=0.5) semi minor axes.
Despite I already know the magnitudes of three axes, I would like to try to use the pca command to find the magnitudes of semi_axes as well as the corresponding vectors.
The ellipsoid stl file in attached.
I use the stlread command to read this simple stl file.
Then perform the pca to the vertices.
[coeff,score,latent,tsquared,explained,mu] =pca(T.Points)
I believe that the ''coeff'' matrix gives me three principal directions (the unit vectors of three semi axes) in terms of column vectors of the ellipsoid.
However, I don't know how to find the ''magnitude of semi axes'' (i.e. max_principal value=1, other tow=0.5) of this ellipsoid from pca data.
After I refer the documentation of pca command, I still fail to find the information could answer my question.
Could anyone give me the hint about it?
Really appreciate for that!!!
Sincerely
Daniel Chou
##### 0 Kommentare-2 ältere Kommentare anzeigen-2 ältere Kommentare ausblenden

Melden Sie sich an, um zu kommentieren.

### Akzeptierte Antwort

Gaurav Garg am 22 Feb. 2021
Hi Daniel,
plot (score(:,1), score(:,2))
On running the above function, you will find a 2-D ellipse which has magnitude same as the magnitude of the original 3-D ellipsoid.
You can also plot the other principal axis by -
plot (score(:,1), score(:,3))
plot (score(:,3), score(:,2))
##### 0 Kommentare-2 ältere Kommentare anzeigen-2 ältere Kommentare ausblenden

Melden Sie sich an, um zu kommentieren.

### Kategorien

Mehr zu Dimensionality Reduction and Feature Extraction finden Sie in Help Center und File Exchange

### Community Treasure Hunt

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

Start Hunting!

Translated by