Plsnipal: Difference between revisions
Jump to navigation
Jump to search
imported>Chuck No edit summary |
imported>Chuck |
||
Line 1: | Line 1: | ||
===Purpose=== | ===Purpose=== | ||
Calculate a single latent variable for partial least squares | Calculate a single latent variable for partial least squares (PLS) regression. | ||
===Synopsis=== | ===Synopsis=== |
Revision as of 21:03, 8 October 2008
Purpose
Calculate a single latent variable for partial least squares (PLS) regression.
Synopsis
- [p,q,w,t,u] = plsnipal(x,y)
Description
PLSNIPAL is called by the routine pls to calculate each latent variable in a partial least squares regression.
Inputs
- x = x-block input data (for first PLS latent variable), or x-block residuals (for subsequent latent variables)
- y = y-block input data (for first PLS latent variable), or y-block residuals (for subsequent latent variables)
Outputs
- p = x-block latent variable loadings
- q = y-block latent variable loadings
- w = x-block latent variable weights
- t = x-block latent variable scores
- u = y-block latent variable scores