File size: 6,727 Bytes
c711202 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 | """
Replot the calibration + scalability figures using a TuePlots theme,
reading saved outputs from `results/` rather than re-running the experiments.
Run from the repo root:
julia --project=. experiments/source_identification/replot_figures.jl
"""
using CairoMakie
using TuePlots
using CSV, DataFrames
using NPZ
const OUT_DIR = get(ENV, "GPFVM_FIGURES_DIR", joinpath(@__DIR__, "figures"))
const CALIB_DIR = joinpath(@__DIR__, "results", "calibration")
const SCAL_DIR = joinpath(@__DIR__, "results", "scalability")
# ──────────────────────────────────────────────────────────────────────────────
# Theme helpers
# ──────────────────────────────────────────────────────────────────────────────
function icml_theme(; single_column=true, nrows=1, ncols=1, ratio=0.75)
Theme(
TuePlots.SETTINGS[:ICML];
font=true, fontsize=true, figsize=true,
single_column=single_column,
nrows=nrows, ncols=ncols,
subplot_height_to_width_ratio=ratio,
)
end
# ──────────────────────────────────────────────────────────────────────────────
# Figure 1: calibration_plot.pdf (Appendix C)
# ──────────────────────────────────────────────────────────────────────────────
function replot_calibration()
df = CSV.read(joinpath(CALIB_DIR, "coverage.csv"), DataFrame)
nominal = df.nominal
emp_s = df.source_coverage
emp_c = df.conc_coverage
set_theme!(icml_theme(single_column=true, nrows=1, ncols=1, ratio=1.0))
fig = Figure()
ax = Axis(fig[1, 1];
xlabel="Nominal coverage",
ylabel="Empirical coverage",
aspect=1,
)
xlims!(ax, 0.45, 1.0); ylims!(ax, 0.45, 1.0)
lines!(ax, [0.45, 1.0], [0.45, 1.0]; color=:gray60, linestyle=:dash, linewidth=1, label="Perfect")
scatterlines!(ax, nominal, emp_s; color=:crimson, linewidth=1.6, markersize=6, label="Source")
scatterlines!(ax, nominal, emp_c; color=:royalblue, linewidth=1.6, markersize=6, label="Concentration")
axislegend(ax; position=:lt, framevisible=false)
out = joinpath(OUT_DIR, "calibration_plot.pdf")
save(out, fig, pt_per_unit=1)
set_theme!()
@info "Wrote $out"
end
# ──────────────────────────────────────────────────────────────────────────────
# Figure 2: spatial_coverage_95.pdf (Appendix C)
# ──────────────────────────────────────────────────────────────────────────────
function replot_spatial_coverage()
d = npzread(joinpath(CALIB_DIR, "calibration_data.npz"))
xs = d["xs"]
ys = d["ys"]
cov_s = d["spatial_s_cov_95"]
cov_c = d["spatial_c_cov_95"]
set_theme!(icml_theme(single_column=true, nrows=1, ncols=2, ratio=1.0))
fig = Figure()
ax1 = Axis(fig[1, 1]; aspect=DataAspect(), title="Source", xlabel=L"x", ylabel=L"y")
ax2 = Axis(fig[1, 2]; aspect=DataAspect(), title="Concentration", xlabel=L"x",
yticklabelsvisible=false, yticksvisible=false)
hm = heatmap!(ax1, xs, ys, cov_s; colormap=:RdBu, colorrange=(0.8, 1.0))
heatmap!(ax2, xs, ys, cov_c; colormap=:RdBu, colorrange=(0.8, 1.0))
Colorbar(fig[1, 3], hm; label="Empirical coverage at 95% nominal", width=8)
out = joinpath(OUT_DIR, "spatial_coverage_95.pdf")
save(out, fig, pt_per_unit=1)
set_theme!()
@info "Wrote $out"
end
# ──────────────────────────────────────────────────────────────────────────────
# Figure 3: scaling_timing_Ns.pdf (Appendix D)
# ──────────────────────────────────────────────────────────────────────────────
function replot_scaling_timing()
df = CSV.read(joinpath(SCAL_DIR, "scalability_results.csv"), DataFrame)
Ns = df.n_total
t_total = df.time_total_s
t_sparse = hasproperty(df, :time_sparse_prec_c_s) ? df.time_sparse_prec_c_s : nothing
t_cond = hasproperty(df, :time_conditioning_s) ? df.time_conditioning_s : nothing
t_post = hasproperty(df, :time_posterior_stats_s) ? df.time_posterior_stats_s : nothing
set_theme!(icml_theme(single_column=true, nrows=1, ncols=1, ratio=0.75))
fig = Figure()
ax = Axis(fig[1, 1];
xlabel=L"Total DOF $N_s$",
ylabel="Wall-clock time (s)",
xscale=log10, yscale=log10,
)
# Reference line O(N^{3/2})
Ns_ref = collect(range(extrema(Ns)...; length=20))
ref_anchor = t_total[1] / (Ns[1]^1.5)
lines!(ax, Ns_ref, ref_anchor .* Ns_ref .^ 1.5;
color=:gray60, linestyle=:dash, linewidth=1, label=L"\mathcal{O}(N_s^{3/2})")
scatterlines!(ax, Ns, t_total; color=:black, linewidth=1.6, markersize=6, label="Total")
t_sparse !== nothing && scatterlines!(ax, Ns, t_sparse; color=:royalblue, linewidth=1.4, markersize=5, label="Sparse prec.")
t_cond !== nothing && scatterlines!(ax, Ns, t_cond; color=:darkorange, linewidth=1.4, markersize=5, label="Conditioning")
t_post !== nothing && scatterlines!(ax, Ns, t_post; color=:purple, linewidth=1.4, markersize=5, label="Posterior stats")
axislegend(ax; position=:lt, framevisible=false)
out = joinpath(OUT_DIR, "scaling_timing_Ns.pdf")
save(out, fig, pt_per_unit=1)
set_theme!()
@info "Wrote $out"
end
# ──────────────────────────────────────────────────────────────────────────────
function main()
isdir(OUT_DIR) || mkpath(OUT_DIR)
replot_calibration()
replot_spatial_coverage()
replot_scaling_timing()
end
main()
|