The Principal component analysis (PCA) node in SciCraft has the
following node arguments:
- Input ports:
X: The input data matrix. Each row is an object and each
column contains the different variables (features)
- Input arguments:
- permutations: Integer specifying the number of permutations of
the input matrix used to estimate the null distribution of the
eigenvalues.
- Ncomp: Integer specifying the number of principal
components wanted. Used in the computation of the residual matrix.
- Output ports:
- ComponentPvalues: Vector that contains the estimated
null distribution of the eigenvalues.
- pcaModel: An R object containing the following
elements:
- center: Vector containing the mean value of each
column of the input matrix
.
- Eigenvalues: Vector containing the eigenvalues of the
estimated covariance matrix.
- LoadingsMatrix: Matrix where column
contains the
loadings of principal component number
, i.e the eigenvectors of
the correlation matrix.
- scale: Vector containing the sample standard
deviation of each column of the input matrix
.
- ScoresMatrix: Matrix containing the scores of the
supplied data onto the principal components.
Bjørn Kåre Alsberg
2006-04-06