textir (version 1.4)

pls: Partial Least Squares

Description

A simple partial least squares procedure.

Usage

pls(F, y, K=1, scale=TRUE, verb=TRUE)

Arguments

F

The covariate matrix, in either simple_triplet_matrix or matrix format.

y

The response vector.

K

The number of desired PLS directions.

scale

An indicator for whether to standardize F; usually a good idea. If scale=TRUE, F will be scaled to have variance-one columns.

verb

Whether or not to print a small progress script.

Value

• A pls object list with the following entries
• y
  The response vector.
• F
  The covariate matrix. If scale=TRUE, scaled to have variance-one columns.
• z
  The pls directions F%*%phi.
• v
  Response factors.
• yhat
  K columns of fitted values for each number of directions.
• fwdmod
  The lm object from forward regression lm(as.numeric(y)~z).
• scale
  Indicator for whether input F was scaled.

Details

Fits the Partial Least Squares algorithm described in Taddy (2011; Section 3.1). In particular, we obtain loadings phi[,k] as the correlation between F and factors v[,k], where v[,1] is initialized at normalize(as.numeric(y)) and subsequent factors are orthogonal to the k'th pls direction, z[,k]=F%*%phi[,k].

References

Taddy (2011), Inverse Regression for Analysis of Sentiment in Text. http://arxiv.org/abs/1012.2098

Wold, H. (1975), Soft modeling by latent variables: The nonlinear iterative partial least squares approach. In Perspectives in Probability and Statistics, Papers in Honour of M.S. Bartlett.

See Also

plot.pls, normalize, freq, corr, we8there, congress109

Examples

data(congress109)
fit <- pls(freq(congress109Counts), congress109Ideology$repshare, K=4)
plot(fit, pch=21, bg=c(4,3,2)[congress109Ideology$party])

data(we8there)
fit <- pls(freq(we8thereCounts),as.factor(we8thereRatings$Overall))
plot(fit, col=c(2,2,2,3,3))