phytoclass (version 2.3.1)

SAALS: Apply the steepest descent algorithm to optimize pigment ratios in phytoplankton classification. Loosely wraps Seepest_Descent function.

Description

Apply the steepest descent algorithm to optimize pigment ratios in phytoplankton classification. Loosely wraps Seepest_Descent function.

Usage

SAALS(Ft, min.value, max.value, place, S, cm, num.loops)

Value

A list containing: 1: The optimized F matrix 2: Final RMSE value

Arguments

Ft

Initial F matrix containing pigment ratios

min.value

UNUSED?

max.value

UNUSED?

place

Vector of indices where F matrix has non-zero values

S

Matrix of sample measurements

cm

Vector of bounded weights for each pigment

num.loops

Maximum number of iterations for the steepest descent

Examples

 Fmat <- as.matrix(phytoclass::Fm)
 S <- as.matrix(phytoclass::Sm)
 S_weights <- as.numeric(phytoclass:::Bounded_weights(S))
 place <- which(Fmat[, seq(ncol(Fmat) - 2)] > 0)
 num.loops <- 2
 # Run SAALS
 result <- phytoclass:::SAALS(Fmat, NULL, NULL, place, S, S_weights, num.loops)