vc_plot_stationary: Inspect two-dimensional stationary candidates and search...

View source: R/plot_stationary.R

vc_plot_stationaryR Documentation

Inspect two-dimensional stationary candidates and search evaluations

Description

Inspect two-dimensional stationary candidates and search evaluations

Usage

vc_plot_stationary(
  result,
  f,
  xlim = NULL,
  ylim = NULL,
  n = 61L,
  run = 1L,
  show_history = TRUE,
  animate = TRUE,
  max_frames = 40L,
  frame_ms = 300,
  colors = c(minimum = "#188367", maximum = "#c44736", saddle = "#8250a3", inconclusive =
    "#dd8800", history = "#24476b"),
  colorscale = "Viridis",
  alpha = 0.6,
  marker_size = 10,
  line_width = 2,
  title = NULL,
  xlab = "x",
  ylab = "y"
)

Arguments

result

Result of vc_stationary_points with two input coordinates.

f

Same scalar field used in the search, accepting one vector.

xlim, ylim

Finite increasing display bounds; NULL uses the search window, or the range of starts/candidates enlarged by one coordinate unit.

n

Integer samples per axis, at least two. Matrix rows represent y and columns x. Nonfinite or unavailable field samples remain gaps.

run

Integer run index whose objective evaluation history is shown.

show_history, animate

Logical switches. Playback shows stored objective evaluations, not optimizer iterations or a Newton/descent path.

max_frames

Positive integer maximum displayed history samples. Deterministic thinning keeps first/last samples; full history stays in result.

frame_ms

Positive playback duration in milliseconds.

colors

Named minimum, maximum, saddle, inconclusive and history colors.

colorscale

Plotly colorscale name or list for the sampled contours.

alpha

Opacity between zero and one for search-evaluation markers.

marker_size

Positive candidate marker size.

line_width

Positive history-line width.

title

Optional scalar title.

xlab, ylab

Scalar axis labels.

Details

No derivative or optimization is recomputed. Only f is sampled for context and checked against stored candidate values. Search evaluations may move uphill and include optimizer probes; their connecting polyline is not a mathematical flow. Failed residual evaluations are omitted from geometry but remain in the calculation history. All candidate labels are conditional numerical evidence; this plot does not establish regularity or completeness.

Value

Plotly contour widget with conditional candidate labels and optional evaluation playback. Hover retains residuals, stopping codes and statuses.

Examples

if(requireNamespace("plotly",quietly=TRUE)) {
  f <- function(v)v[1]^2+2*v[2]^2
  r <- vc_stationary_points(f,c(.1,.2),assume_c2=TRUE)
  r |> vc_plot_stationary(f=f)
}

vectorialcalculus documentation built on Nov. 17, 2026, 1:09 a.m.