GitHub

Plotting options

Most keywords from Plots.jl are supported. Furthermore, the following options can be used to change the appearance of gird plots (using the grids from Plotting basics for the examples):

Fill colors

The fill color of each interval/block is determined from a color function that can be specified with the fill_color_function keyword. To illustrate the principle, consider the default color function:

function default_color_function(current_weight,grid_weights)
	# Get the weight range.
	min_w, max_w  = extrema(grid_weights)
	if min_w == max_w
		opacity = 1
	else
		# Calculate the proportion of the current_weight.
		opacity = (current_weight - min_w)/(max_w - min_w)
	end
	return (opacity,"Midnight Blue")
end

The color function will be applied internally to every interval/block and must accept two arguments: (val,Arr), where val will be the weight of the current interval/block and where Arr will be the vector of all wights in the grid. The color function must return two arguments, an opacity (number between 0 and 1) and a color compatible with Plots.jl.

As example, we can implement the color gradient :thermal form ColorsSchemes.jl:

using ColorSchemes

thermal = cgrad(:thermal)

function thermal_function(current_weight,grid_weights)
	min_w, max_w  = extrema(grid_weights)
	weight_proportion = (current_weight - min_w)/(max_w-min_w)
	return (1,thermal[weight_proportion])
end

To apply the new color function, we set fill_color_function = thermal_function for the plotting:

plot(two_dim_grid, fill_color_function = thermal_function)
Example block output

At the moment, there is no way to add automatically generated colorbars to the plot. As workaround, the colorbars can be created manually, e.g. as explained in Colorbars

The color function can also be used for a OneDimGrid:

plot(one_dim_grid, fill_color_function = thermal_function)
Example block output

Since the bar height for OneDimGrid-plots already depicts the weight for each interval, a single fill color can be sufficient: 

plot(one_dim_grid, fill_color_function = (w,W) -> (1,"Midnight Blue"))
Example block output

Line colors

The color of the lines that separate the intervals/blocks can be changed with the line_color_function keyword. It works in the same way as fill_color_function:

plot(two_dim_grid, line_color_function = (w,W) ->  (1,"dark red"))
Example block output
plot(one_dim_grid, line_color_function = (w,W) ->  (1,"dark red"))
Example block output

It is possible to pass the same function to both line_color_function and fill_color_function. This allows to hide the lines that separate the intervals/blocks. 

plot(one_dim_grid, line_color_function = (w,W) -> (1,"Midnight Blue"),
	fill_color_function = (w,W) -> (1,"Midnight Blue"))
Example block output

Since the intervals/blocks are plotted one after another, interval/block-specific line colors can lead to visual artifacts. This effect can be exaggerated with increased line widths and the thermal_function that was defined in Fill colors:

plot(two_dim_grid,linewidth = 6, line_color_function = thermal_function,
	fill_color_function = thermal_function)
Example block output

Reducing the line width, on the other hand, can fix the visual artifacts

plot(two_dim_grid, linewidth = 0.5, line_color_function = thermal_function,
	fill_color_function = thermal_function)
Example block output

OneDimGrid bar height

The default plotting behavior for OneDimGrid-plots and Grid-plots differs, since one-dimensional grids can use the free y-axis to depict the weights of intervals. This behavior can be changed with the height_function that determines the bar-height, based on the weight of the corresponding interval. The default function is x->x, which produces bar heights equal to the respective interval weights.

To obtain OneDimGrid-plots that behave like Grid-plots, a constant height value needs to be set:

plot(one_dim_grid, height_function = x->1)
Example block output

The height_function allows to separate the bar heights from the actual gird weights. For example, the absolute value can be plotted, but the true weight value can be used for the fill colors (using the thermal_function defined in Fill colors):

plot(one_dim_grid, height_function = x->abs(x), fill_color_function = thermal_function)
Example block output

OneDimGrid along the y-axis

A OneDimGrid can be plotted along the y-axis by setting vertical=true:

plot(one_dim_grid, vertical =true)
Example block output

Colorbars

Currently, there is no way to create automatic colorbars. As workaround, a colorbar can be created manually with a OneDimGrid-plot. For the example, we use the 2-dim example grid two_dim_grid and the thermal_function that was defined in Fill colors.

First, we extract the weight range from two_dim_grid.

weight_range = extrema(export_weights(two_dim_grid))
(0.006819348298638676, 1.9931806517013615)

Then, we create a new 1-dim grid color_grid that has weight_range as domain, and set up 100 intervals to get a continuous color bar:

# 101 discretization points -> 100 intervals
color_grid = create_grid(LinRange(weight_range...,101))

Next, we import 100 increasing, linearly spaced wights (e.g. LinRange(0,1,100)):

import_weights!(color_grid, LinRange(0,1,100))

To plot the colorbar, we use the following options

  • thermal_function for line_color_function and fill_color_function
  • height_function = x-> 1 to obtain a bar
  • yaxis=false to disable the y-axis.
plot(color_grid, fill_color_function = thermal_function,
	line_color_function = thermal_function,
	height_function = x-> 1, yaxis = false)
Example block output

In this form, the colorbar is not very useful. To plot the colorbar next to the two_dim_grid plot, we can use subplots and layouts:

p1 = plot(two_dim_grid, fill_color_function = thermal_function)
p2 = plot(color_grid, fill_color_function = thermal_function,
	line_color_function = thermal_function,
	height_function = x-> 1, size = (600,50), yaxis = false)
plot(p1,p2, layout = @layout [a{0.9h}; b])
Example block output

In the same way, a vertical colorbar can be created (using Measures.jl to adjust the margins):

using Measures
p1 = plot(two_dim_grid, fill_color_function = thermal_function)
p2 = plot(color_grid, fill_color_function = thermal_function,
	line_color_function = thermal_function,
	height_function = x-> 1, vertical = true,
	xaxis = false, left_margin=5mm)
plot(p1,p2, layout = @layout [a{0.9w} b])
Example block output