plotConvergence: plotConvergence

In mrparker909/optimizeAPA: An R Package for Arbitrary Precision Optimization of Multi-parameter Functions

Description

Plot the path to convergence of an optimization algorithm (such as optim_DFP_APA).

Usage

plotConvergence(
  optimOutput,
  variable = NULL,
  digits = 5,
  labels = T,
  flip_axes = F
)

Arguments

optimOutput	The result of running an optimization algorithm with keepValues=TRUE
variable	Default is NULL, will print convergence plot for all variables, otherwise an integer vector (such as c(1,3)) indicating which variables to print convergence plot for.
digits	Default is 5. Number of digits to print on convergence plot.
labels	If TRUE, print function values as labels on the plot.
flip_axes	If TRUE, will plot the convergence horizontally rather than vertically.

Examples

# N dimensional quadratic
funcND <- function(par, center, precBits=64) {
  par <- Rmpfr::mpfr(par, precBits)
  log(sum(exp((par-center)^2)))
} 
opt <- optimizeAPA::optim_DFP_APA(starts = c(0,0,0), func = funcND, center=c(1,2,3), keepValues=TRUE)
plotConvergence(opt, variable=c(1,3), labels=F)

mrparker909/optimizeAPA documentation built on June 28, 2020, 3:57 p.m.