Hi Matlabman,
In MATLAB, the ‘pca’ function performs Principal Component Analysis (PCA) by default without applying any rotation techniques such as varimax or oblique rotation. PCA in MATLAB is designed to transform the data into a new coordinate system where the greatest variances by any projection of the data come to lie on the first coordinates (called principal components), the second greatest variances on the second coordinates, and so on. This transformation is achieved through eigenvalue decomposition of the data's covariance matrix or through singular value decomposition (SVD) of the data matrix.
Since PCA focuses on maximizing variance and maintaining orthogonality, it does not inherently involve any rotation techniques like varimax or oblique rotations, which are often used in factor analysis to make the output more interpretable by rotating the factors to achieve a simpler and more meaningful structure.
If you need to apply rotation techniques such as varimax or oblique rotation after performing PCA, you will typically need to do this as a post-processing step. MATLAB provides functions such as ‘rotatefactors’ for this purpose. You can first extract the principal components using the ‘pca’ function and then apply a rotation to the factor loadings obtained from the principal components. This approach allows you to benefit from the variance-maximizing property of PCA while also achieving a rotated solution that may be easier to interpret.
For more details on the function refer the following documentation links, refer the following links:
- https://www.mathworks.com/help/releases/R2019a/stats/rotatefactors.html
- https://www.mathworks.com/help/releases/R2019a/stats/pca.html
Hope this helps.
