plsDA: PLS Discriminant Analysis

In DiscriMiner: Tools of the Trade for Discriminant Analysis

Description

Performs a Partial Least Squares (PLS) Discriminant Analysis by giving the option to include a random leave-k fold out cross validation

Usage

  plsDA(variables, group, autosel = TRUE, comps = 2,
    validation = NULL, learn = NULL, test = NULL,
    cv = "LOO", k = NULL, retain.models = FALSE)

Arguments

variables	matrix or data frame with explanatory variables
group	vector or factor with group memberships
autosel	logical indicating automatic selection of PLS components by cross-validation. Default autosel=TRUE
comps	integer greater than one indicating the number of PLS components to retain. Used only when autosel=FALSE
validation	type of validation, either NULL or "learntest". Default NULL
learn	optional vector of indices for a learn-set. Only used when validation="learntest". Default NULL
test	optional vector of indices for a test-set. Only used when validation="learntest". Default NULL
cv	string indicating the type of crossvalidation. Avialable options are "LOO" (Leave-One-Out) and "LKO" (Leave-K fold-Out)
k	fold left out if using LKO (usually 7 or 10)
retain.models	whether to retain lower models (i.e. all lower component results)

Details

When validation=NULL leave-one-out (loo) cross-validation is performed.
When validation="learntest" validation is performed by providing a learn-set and a test-set of observations.

Value

An object of class "plsda", basically a list with the following elements:

functions	table with discriminant functions
confusion	confusion matrix
scores	discriminant scores for each observation
loadings	loadings
y.loadings	y loadings
classification	assigned class
error_rate	misclassification error rate
components	PLS components
Q2	quality of loo cross-validation
R2	R-squared coefficients
VIP	Variable Importance for Projection
comp_vars	correlations between components and variables
comp_group	correlations between components and groups

Author(s)

Charles Determan Jr, Gaston Sanchez

References

Tenenhaus M. (1998) La Regression PLS. Editions Technip, Paris.

Perez-Enciso M., Tenenhaus M. (2003) Prediction of clinical outcome with microarray data: a partial least squares discriminant analysis (PLS-DA) approach. Human Genetics 112: 581-592.

See Also

classify, geoDA, linDA, quaDA

Examples

## Not run: 
# load iris dataset
data(iris)

# PLS discriminant analysis specifying number of components = 2
my_pls1 = plsDA(iris[,1:4], iris$Species, autosel=FALSE, comps=2)
my_pls1$confusion
my_pls1$error_rate
# plot circle of correlations
plot(my_pls1)

# PLS discriminant analysis with automatic selection of components
my_pls2 = plsDA(iris[,1:4], iris$Species, autosel=TRUE)
my_pls2$confusion
my_pls2$error_rate

# linear discriminant analysis with learn-test validation
learning = c(1:40, 51:90, 101:140)
testing = c(41:50, 91:100, 141:150)
my_pls3 = plsDA(iris[,1:4], iris$Species, validation="learntest",
learn=learning, test=testing)
my_pls3$confusion
my_pls3$error_rate

## End(Not run)

Example output

            predicted
original     setosa versicolor virginica
  setosa         49          1         0
  versicolor      0         30        20
  virginica       0          7        43
[1] 0.1866667
            predicted
original     setosa versicolor virginica
  setosa         49          1         0
  versicolor      0         30        20
  virginica       0          7        43
[1] 0.1866667
            predicted
original     setosa versicolor virginica
  setosa          9          1         0
  versicolor      0          8         2
  virginica       0          2         8
[1] 0.1666667

DiscriMiner documentation built on May 1, 2019, 10:32 p.m.