ggplotHeatmap: Visualize Heatmap of Multi-Objective Gradients.

Description Usage Arguments Value Examples

View source: R/ggplotHeatmap.R

Description

Visualization of the multi-objective gradient landscape by means of a heatmap and on the basis of the ggplot2-package.

Usage

1
2
ggplotHeatmap(df, log.scale = TRUE, impute.zero = TRUE,
  minimalistic.image = FALSE, color.palette, legend.position, ...)

Arguments

df

[data.frame]
Data frame as returned by computeCumulatedPathLengths.

log.scale

[logical(1L)]
Should the resulting heights be displayed on a log-scale? The default is TRUE.

impute.zero

[logical(1L)]
Should height values, which are exactly zero be imputed by a value half the magnitude of the smallest non-zero height? Otherwise ggplot will automatically color the corresponding tiles by a color representing NA values (usually grey). Note that this parameter is only relevant in case of log.scale = TRUE. The default is TRUE.

minimalistic.image

[logical(1L)]
Should all information surrounding the image (axes, legends, background, etc.) be discarded? The default is FALSE.

color.palette

[character]
Vector of colors used for visualizing the different heights of the landscape. By default, this function tries to use the color palettes from fields::tim.color or viridis. However, if neither of these packages is installed, it will use terrain.colors.

legend.position

[character(1L)]
On which side of the plot, should the legend be located? If this information is not provided and minimalisitic.image = FALSE, the legend will be placed on the right side.

...

[any]
Further arguments to be passed to the geom_tile function of ggplot.

Value

[ggplot]
A ggplot object displaying the multi-objective gradient landscape.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
# Define two single-objective test problems and a grid of points:
fn1 = function(x) sum((x - c(0.2, 1))^2)
fn2 = function(x) sum((x - c(0.5, 0.5))^2)
points = as.matrix(expand.grid(x1 = seq(0, 0.7, 0.005), x2 = seq(0, 1.25, 0.005)))

# Compute the corresponding gradients and the cumulated path lengths:
gradients = computeGradientField(points, fn1, fn2)
x = computeCumulatedPathLengths(points, gradients)

# Finally, we can visualize the resulting multi-objective "landscape":
ggplotHeatmap(x)

# If one is only interested in the image itself, but not in any labels, legends, etc.
# one should set minimalistic.image = TRUE:
ggplotHeatmap(x, minimalistic.image = TRUE)

kerschke/mogsa documentation built on Oct. 27, 2018, 12:13 a.m.