I there any built in tool in MATLAB that perform Principle Component Analysis?

Hi

I am fairly new with PCA and Matlab.

I am trying to perform principle component analysis to reduce a data set that have 18 features to the 2 or 3 principle components and then using them to build clusters to classify several faults in a system. I have 270 samples for each feature. I have 10 faults, each 27 samples represent a specific fault in the system.

I am planning to train and test the algorithm to make sure that my algorithm can identify theses faults. It is very important for me that I get all the factors that are used in the linear correlation to calculate the first, second and third principle components so that after building my cluster I can use them to calculate the principle components when receiving the 18 features from the system then see in which cluster they belong.

I am trying write a code that calculates all the matrices from scratch to do PCA, but if there is any built in tool in Matlab for PCA, it will save me a lot of time.

Is there any? and if so is there any tutorials on how to use it?

Thanks in advance.

Answer 1

[Thorsten]

Matlab's PCA is called

princomp

You can find out using

lookfor principal

Which results in my system in

pcacov                         - Principal Components Analysis (PCA) using a covariance matrix.
pcares                         - Residuals from a Principal Components Analysis (PCA).
princomp                       - Principal Components Analysis (PCA) from raw data.
orthoregdemo                   - Fitting an Orthogonal Regression Using Principal Components Analysis
pcrsse                         - SSE for Principal Components Regression cross-validation.
plspcrdemo                     - Partial Least Squares Regression and Principal Components Regression

[M.Abuasbeh]

Thank you very much Thorsten!