NMRWish Built-In TCL Functions
pcaNMR: Principal Component Analysis matrix decomposition on spectral data in a region of interest (ROI). See also: readROI, reconNMR.

PCA Decomposition and Reconstruction:

Flag	Argument	Default	Description
`-x`	`xROI`	`0`	ROI ID For Data Matrix.
`-t`	`tROI`	`0`	ROI ID For Score (X-Axis) Vectors.
`-p`	`pROI`	`0`	ROI ID For Loading (Y-Axis) Vectors.
`-q`	`qROI`	`0`	ROI ID For Loading (Z-Axis) Vectors.

PCA Arguments:

`-nc`	`NC`	`4`	Number of Components to Find.
`-itmax`	`iMax`	`4`	Max Iterations Per Component.
`-i`			iList List of Component IDs for Recon.

Notes:
1. If NC is not zero, input matrix X is decomposed
into output T, P, and (for 3D X) Q.
On return, X is replaced with the residual.
2. If NC is zero, the reconstruction formed by
input T, P, and (for 3D X) Q is added into
input matrix X.
3. 2D Case:
X Matrix is xSize*ySize.
T Matrix is xSize*NC.
P Matrix is ySize*NC.
4. 3D Case:
X Matrix is xSize*ySize*zSize.
T Matrix is xSize*NC.
P Matrix is ySize*NC.
Q Matrix is zSize*NC.

Frank Delaglio, Ph.D.