**BREAKING** change of sampleplot behaviour and defaults (#213)

This PR makes two breaking changes to the `sampleplot` plotting utility:

1. **Most default settings have been removed.**
   - `sampleplot` no longer imposes a default seriescolor.
     **How to obtain the previous color:** manually pass `seriescolor = "red"`.
   - The markers have been removed as they seemed more confusing than helpful (overwhelmingly large in the legend, almost invisibly tiny on the plot). The `linealpha` default has been increased to 0.35 to keep visual appearance the same.
     **How to obtain the previous markers:** If you actually liked the tiny markers, [check the diff](https://github.com/JuliaGaussianProcesses/AbstractGPs.jl/pull/213/files#diff-59728b925dcbfab15dd2def5624b3bd8b3a6a609570dc56379f69c8ee0bdef92L124-L127) for the previous default settings that you now need to pass manually.
2. **Multiple samples (`samples` kwarg > 1) are now plotted as a single series.** This means, for example, that passing a string as `label` only adds one legend entry for all samples, not one _per_ sample.
   **How to obtain previous behaviour:** If you actually want each sample to be a different series (for example, to give them different colors), you now need to call `sampleplot` multiple times instead.

Additionally, this PR contains the following non-breaking minor changes:
- minor fix: error message for check of `ribbon_scale` in `plot` recipe
- internal code cleanup (`pop!` for custom-defined kwarg)

Project.toml

@@ -1,7 +1,7 @@
 name = "AbstractGPs"
 uuid = "99985d1d-32ba-4be9-9821-2ec096f28918"
 authors = ["JuliaGaussianProcesses Team"]
-version = "0.4.0"
+version = "0.5.0"
 
 [deps]
 ChainRulesCore = "d360d2e6-b24c-11e9-a2a3-2a2ae2dbcce4"

docs/Project.toml

@@ -4,5 +4,5 @@ Documenter = "e30172f5-a6a5-5a46-863b-614d45cd2de4"
 Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
 
 [compat]
-AbstractGPs = "0.4"
+AbstractGPs = "0.4, 0.5"
 Documenter = "0.27"

examples/regression-1d/Project.toml

@@ -13,7 +13,7 @@ Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
 StatsFuns = "4c63d2b9-4356-54db-8cca-17b64c39e42c"
 
 [compat]
-AbstractGPs = "0.4"
+AbstractGPs = "0.4, 0.5"
 AdvancedHMC = "0.2"
 Distributions = "0.19, 0.20, 0.21, 0.22, 0.23, 0.24, 0.25"
 DynamicHMC = "2.2, 3.1"

examples/regression-1d/script.jl

@@ -212,19 +212,19 @@ mean(logpdf(gp_posterior(x_train, y_train, p)(x_test), y_test) for p in samples)
 # We sample 5 functions from each posterior GP given by the final 100 samples of kernel
 # parameters.
 
-plt = scatter(
-    x_train,
-    y_train;
-    xlim=(0, 1),
-    xlabel="x",
-    ylabel="y",
-    title="posterior (AdvancedHMC)",
-    label="Train Data",
-)
-for p in samples[(end - 100):end]
-    sampleplot!(plt, 0:0.02:1, gp_posterior(x_train, y_train, p); samples=5)
+plt = plot(; xlim=(0, 1), xlabel="x", ylabel="y", title="posterior (AdvancedHMC)")
+for (i, p) in enumerate(samples[(end - 100):end])
+    sampleplot!(
+        plt,
+        0:0.02:1,
+        gp_posterior(x_train, y_train, p);
+        samples=5,
+        seriescolor="red",
+        label=(i == 1 ? "samples" : nothing),
+    )
 end
-scatter!(plt, x_test, y_test; label="Test Data")
+scatter!(plt, x_train, y_train; label="Train Data", markercolor=1)
+scatter!(plt, x_test, y_test; label="Test Data", markercolor=2)
 plt
 
 # #### DynamicHMC
@@ -290,18 +290,11 @@ mean(logpdf(gp_posterior(x_train, y_train, p)(x_test), y_test) for p in samples)
 # We sample a function from the posterior GP for the final 100 samples of kernel
 # parameters.
 
-plt = scatter(
-    x_train,
-    y_train;
-    xlim=(0, 1),
-    xlabel="x",
-    ylabel="y",
-    title="posterior (DynamicHMC)",
-    label="Train Data",
-)
+plt = plot(; xlim=(0, 1), xlabel="x", ylabel="y", title="posterior (DynamicHMC)")
+scatter!(plt, x_train, y_train; label="Train Data")
 scatter!(plt, x_test, y_test; label="Test Data")
 for p in samples[(end - 100):end]
-    sampleplot!(plt, 0:0.02:1, gp_posterior(x_train, y_train, p))
+    sampleplot!(plt, 0:0.02:1, gp_posterior(x_train, y_train, p); seriescolor="red")
 end
 plt
 
@@ -349,18 +342,13 @@ mean(logpdf(gp_posterior(x_train, y_train, p)(x_test), y_test) for p in samples)
 # We sample a function from the posterior GP for the final 100 samples of kernel
 # parameters.
 
-plt = scatter(
-    x_train,
-    y_train;
-    xlim=(0, 1),
-    xlabel="x",
-    ylabel="y",
-    title="posterior (EllipticalSliceSampling)",
-    label="Train Data",
+plt = plot(;
+    xlim=(0, 1), xlabel="x", ylabel="y", title="posterior (EllipticalSliceSampling)"
 )
+scatter!(plt, x_train, y_train; label="Train Data")
 scatter!(plt, x_test, y_test; label="Test Data")
 for p in samples[(end - 100):end]
-    sampleplot!(plt, 0:0.02:1, gp_posterior(x_train, y_train, p))
+    sampleplot!(plt, 0:0.02:1, gp_posterior(x_train, y_train, p); seriescolor="red")
 end
 plt
 

src/util/plotting.jl

@@ -4,7 +4,7 @@
     length(x) == length(gp.x) ||
         throw(DimensionMismatch("length of `x` and `gp.x` has to be equal"))
     scale::Float64 = pop!(plotattributes, :ribbon_scale, 1.0)
-    scale > 0.0 || error("`bandwidth` keyword argument must be non-negative")
+    scale >= 0.0 || error("`ribbon_scale` keyword argument must be non-negative")
 
     # compute marginals
     μ, σ2 = mean_and_var(gp)
@@ -82,16 +82,19 @@ Plot samples from the projection `f` of a Gaussian process versus `x`.
     Make sure to load [Plots.jl](https://github.com/JuliaPlots/Plots.jl) before you use
     this function.
 
+When plotting multiple samples, these are treated as a _single_ series (i.e.,
+only a single entry will be added to the legend when providing a `label`).
+
 # Example
 
 ```julia
 using Plots
 
 gp = GP(SqExponentialKernel())
-sampleplot(gp(rand(5)); samples=10, markersize=5)
+sampleplot(gp(rand(5)); samples=10, linealpha=1.0)
 ```
-The given example plots 10 samples from the projection of the GP `gp`. The `markersize` is modified
-from default of 0.5 to 5.
+The given example plots 10 samples from the projection of the GP `gp`.
+The `linealpha` is modified from default of 0.35 to 1.
 
 ---
     sampleplot(x::AbstractVector, gp::AbstractGP; samples=1, kwargs...)
@@ -115,18 +118,15 @@ SamplePlot((f,)::Tuple{<:FiniteGP}) = SamplePlot((f.x, f))
 SamplePlot((x, gp)::Tuple{<:AbstractVector,<:AbstractGP}) = SamplePlot((gp(x, 1e-9),))
 
 @recipe function f(sp::SamplePlot)
-    nsamples::Int = get(plotattributes, :samples, 1)
+    nsamples::Int = pop!(plotattributes, :samples, 1)
     samples = rand(sp.f, nsamples)
 
+    flat_x = repeat(vcat(sp.x, NaN), nsamples)
+    flat_f = vec(vcat(samples, fill(NaN, 1, nsamples)))
+
     # Set default attributes
-    seriestype --> :line
-    linealpha --> 0.2
-    markershape --> :circle
-    markerstrokewidth --> 0.0
-    markersize --> 0.5
-    markeralpha --> 0.3
-    seriescolor --> "red"
+    linealpha --> 0.35
     label --> ""
 
-    return sp.x, samples
+    return flat_x, flat_f
 end

test/deprecations.jl

@@ -4,11 +4,11 @@
     gp = f(x, 0.1)
 
     plt = @test_deprecated sampleplot(gp, 10)
-    @test plt.n == 10
+    @test plt.n == 1
 
     @test_deprecated sampleplot!(gp, 4)
-    @test plt.n == 14
+    @test plt.n == 2
 
     @test_deprecated sampleplot!(Plots.current(), gp, 3)
-    @test plt.n == 17
+    @test plt.n == 3
 end

test/util/plotting.jl

@@ -6,18 +6,20 @@
     z = rand(10)
     plt1 = sampleplot(z, gp)
     @test plt1.n == 1
-    @test plt1.series_list[1].plotattributes[:x] == sort(z)
+    @test isequal(plt1.series_list[1].plotattributes[:x], vcat(z, NaN))
 
-    plt2 = sampleplot(gp; samples=10)
-    @test plt2.n == 10
-    sort_x = sort(x)
-    @test all(series.plotattributes[:x] == sort_x for series in plt2.series_list)
+    plt2 = sampleplot(gp; samples=3)
+    @test plt2.n == 1
+    plt2_x = plt2.series_list[1].plotattributes[:x]
+    plt2_y = plt2.series_list[1].plotattributes[:y]
+    @test isequal(plt2_x, vcat(x, NaN, x, NaN, x, NaN))
+    @test length(plt2_y) == length(plt2_x)
+    @test isnan(plt2_y[length(z) + 1]) && isnan(plt2_y[2length(z) + 2])
 
-    z = rand(7)
-    plt3 = sampleplot(z, f; samples=8)
-    @test plt3.n == 8
-    sort_z = sort(z)
-    @test all(series.plotattributes[:x] == sort_z for series in plt3.series_list)
+    z3 = rand(7)
+    plt3 = sampleplot(z3, f; samples=2)
+    @test plt3.n == 1
+    @test isequal(plt3.series_list[1].plotattributes[:x], vcat(z3, NaN, z3, NaN))
 
     # Check recipe dispatches for `FiniteGP`s
     rec = RecipesBase.apply_recipe(Dict{Symbol,Any}(), gp)