simulator_plot: Plot simulator outputs for multiple waves

In hmer: History Matching and Emulation Package

Plot simulator outputs for multiple waves

Description

Plots the simulator results for points at successive waves.

Usage

simulator_plot(
  wave_points,
  z,
  zero_in = TRUE,
  palette = NULL,
  wave_numbers = seq(ifelse(zero_in, 0, 1), length(wave_points) - ifelse(zero_in, 1, 0)),
  normalize = FALSE,
  logscale = FALSE,
  byhit = FALSE,
  barcol = "#444444",
  ...
)

Arguments

wave_points	The set of wave points, as a list of data.frames
z	The set of target values for each output
zero_in	Is wave zero included? Default: TRUE
palette	If a larger palette is required, it should be supplied here.
wave_numbers	Which waves to plot. If not supplied, all waves are plotted.
normalize	If true, plotting is done with rescaled target bounds.
logscale	If true, targets are log-scaled before plotting.
byhit	Should runs be grouped by number of targets hit, rather than wave?
barcol	The colour of the target error bars/bounds
...	Optional parameters (not to be used directly)

Details

The values plotted are the outputs from the simulator; the points passed to it are the points suggested by that wave of emulators. By default, wave 0 is included. A colour scheme is chosen outright for all invocations of this function: it is a 10-colour palette. If more waves are required, then an alternative palette should be selected.

The output can be plotted in a number of ways: raw; with outputs transformed to log scale; or with targets normalised so that target bounds are all [-1, 1]. These two options may be helpful in visualising behaviour when outputs have vastly different scales, but one still wishes to see them all in the same plot: these options can be toggled by setting logscale = TRUE or normalize = TRUE respectively. The data can be grouped in two ways, either colouring by wave of emulation (default) or by the number of targets hit; the latter option is enabled by setting byhit = TRUE.

Value

A ggplot object.

See Also

Other visualisation tools: behaviour_plot(), diagnostic_wrap(), effect_strength(), emulator_plot(), hit_by_wave(), output_plot(), plot_actives(), plot_lattice(), plot_wrap(), space_removed(), validation_pairs(), wave_dependencies(), wave_points(), wave_values()

Examples

 simulator_plot(SIRMultiWaveData, SIREmulators$targets)
 simulator_plot(SIRMultiWaveData[2:4], SIREmulators$targets,
  zero_in = FALSE, wave_numbers = c(1,3))
 simulator_plot(SIRMultiWaveData, SIREmulators$targets, byhit = TRUE)

hmer documentation built on June 22, 2024, 9:22 a.m.