heatmap(x, y, values)

Plots a heatmap as an image on x, y (defaults to interpretation as dimensions).


Specific to Heatmap

  • interpolate::Bool = false sets whether colors should be interpolated.

Color attributes

  • colormap::Union{Symbol, Vector{<:Colorant}} = :viridis sets the colormap that is sampled for numeric colors. PlotUtils.cgrad(...), Makie.Reverse(any_colormap) can be used as well, or any symbol from ColorBrewer or PlotUtils. To see all available color gradients, you can call Makie.available_gradients().

  • colorscale::Function = identity color transform function. Can be any function, but only works well together with Colorbar for identity, log, log2, log10, sqrt, logit, Makie.pseudolog10 and Makie.Symlog10.

  • colorrange::Tuple{<:Real, <:Real} sets the values representing the start and end points of colormap.

  • nan_color::Union{Symbol, <:Colorant} = RGBAf(0,0,0,0) sets a replacement color for color = NaN.

  • lowclip::Union{Nothing, Symbol, <:Colorant} = nothing sets a color for any value below the colorrange.

  • highclip::Union{Nothing, Symbol, <:Colorant} = nothing sets a color for any value above the colorrange.

  • alpha = 1.0 sets the alpha value of the colormap or color attribute. Multiple alphas like in plot(alpha=0.2, color=(:red, 0.5), will get multiplied.

Generic attributes

  • visible::Bool = true sets whether the plot will be rendered or not.

  • overdraw::Bool = false sets whether the plot will draw over other plots. This specifically means ignoring depth checks in GL backends.

  • transparency::Bool = false adjusts how the plot deals with transparency. In GLMakie transparency = true results in using Order Independent Transparency.

  • fxaa::Bool = true adjusts whether the plot is rendered with fxaa (anti-aliasing).

  • inspectable::Bool = true sets whether this plot should be seen by DataInspector.

  • depth_shift::Float32 = 0f0 adjusts the depth value of a plot after all other transformations, i.e. in clip space, where 0 <= depth <= 1. This only applies to GLMakie and WGLMakie and can be used to adjust render order (like a tunable overdraw).

  • model::Makie.Mat4f sets a model matrix for the plot. This replaces adjustments made with translate!, rotate! and scale!.

  • space::Symbol = :data sets the transformation space for box encompassing the volume plot. See Makie.spaces() for possible inputs.


Two vectors and a matrix

using CairoMakie

xs = range(0, 10, length = 25)
ys = range(0, 15, length = 25)
zs = [cos(x) * sin(y) for x in xs, y in ys]

heatmap(xs, ys, zs)

Two ranges and a function

using CairoMakie

function mandelbrot(x, y)
    z = c = x + y*im
    for i in 1:30.0; abs(z) > 2 && return i; z = z^2 + c; end; 0

heatmap(-2:0.001:1, -1.1:0.001:1.1, mandelbrot,
    colormap = Reverse(:deep))

Three vectors

There must be no duplicate combinations of x and y, but it is allowed to leave out values.

using CairoMakie

xs = [1, 2, 3, 1, 2, 3, 1, 2, 3]
ys = [1, 1, 1, 2, 2, 2, 3, 3, 3]
zs = [1, 2, 3, 4, 5, 6, 7, 8, NaN]

heatmap(xs, ys, zs)

Colorbar for single heatmap

To get a scale for what the colors represent, add a colorbar. The colorbar is placed within the figure in the first argument, and the scale and colormap can be conveniently set by passing the relevant heatmap to it.

using CairoMakie

xs = range(0, 2π, length=100)
ys = range(0, 2π, length=100)
zs = [sin(x*y) for x in xs, y in ys]

fig, ax, hm = heatmap(xs, ys, zs)
Colorbar(fig[:, end+1], hm)


Colorbar for multiple heatmaps

When there are several heatmaps in a single figure, it can be useful to have a single colorbar represent all of them. It is important to then have synchronized scales and colormaps for the heatmaps and colorbar. This is done by setting the colorrange explicitly, so that it is independent of the data shown by that particular heatmap.

Since the heatmaps in the example below have the same colorrange and colormap, any of them can be passed to Colorbar to give the colorbar the same attributes. Alternativly, the colorbar attributes can be set explicitly.

using CairoMakie

xs = range(0, 2π, length=100)
ys = range(0, 2π, length=100)
zs1 = [sin(x*y) for x in xs, y in ys]
zs2 = [2sin(x*y) for x in xs, y in ys]

joint_limits = (-2, 2)  # here we pick the limits manually for simplicity instead of computing them

fig, ax1, hm1 = heatmap(xs, ys, zs1,  colorrange = joint_limits)
ax2, hm2 = heatmap(fig[1, end+1], xs, ys, zs2, colorrange = joint_limits)

Colorbar(fig[:, end+1], hm1)                     # These three
Colorbar(fig[:, end+1], hm2)                     # colorbars are
Colorbar(fig[:, end+1], colorrange = joint_limits)  # equivalent


Using a custom colorscale

One can define a custom (color)scale using the ReversibleScale type. When the transformation is simple enough (log, sqrt, ...), the inverse transform is automatically deduced.

using CairoMakie

x = 10.0.^(1:0.1:4)
y = 1.0:0.1:5.0
z = broadcast((x, y) -> x - 10, x, y')

scale = ReversibleScale(x -> asinh(x / 2) / log(10), x -> 2sinh(log(10) * x))
fig, ax, hm = heatmap(x, y, z; colorscale = scale, axis = (; xscale = scale))
Colorbar(fig[1, 2], hm)