fix e2e test

calibration/test/e2e_test.jl

@@ -8,7 +8,7 @@ import EnsembleKalmanProcesses as EKP
 import Statistics: var, mean
 using Test
 
-experiment_dir = joinpath(pkgdir(CA), "test", "calibration")
+experiment_dir = joinpath(pkgdir(CA), "calibration", "test")
 model_interface = joinpath(pkgdir(CA), "calibration", "model_interface.jl")
 include(model_interface)
 
@@ -34,8 +34,9 @@ end
 
 function process_member_data(simdir::SimDir)
     isempty(simdir.vars) && return NaN
-    rsut = get(simdir; short_name = "rsut", reduction = "average", period = "30d")
-    return slice(Cl(rsut); time = 30).data
+    rsut =
+        get(simdir; short_name = "rsut", reduction = "average", period = "30d")
+    return slice(average_xy(rsut); time = 30).data
 end
 
 # Generate observations
@@ -56,7 +57,7 @@ noise = 0.1 * I
 n_iterations = 3
 ensemble_size = 10
 prior = CAL.get_prior(joinpath(experiment_dir, "prior.toml"))
-output_dir = joinpath("output", "single_column_held_suarez_rhoe_equilmoist")
+output_dir = "calibration_end_to_end_test"
 experiment_config = CAL.ExperimentConfig(;
     n_iterations,
     ensemble_size,
@@ -85,18 +86,19 @@ for i in 0:(n_iterations - 1)
     end
     G_ensemble = CAL.observation_map(i)
     CAL.save_G_ensemble(experiment_config, i, G_ensemble)
-    eki = CAL.update_ensemble(experiment_config, i)
+    global eki = CAL.update_ensemble(experiment_config, i)
 end
 
-@testset "Pure Julia Calibration" begin 
+@testset "Pure Julia Calibration" begin
     minimal_eki_test(eki)
 end
 
 backend = CAL.get_backend()
 if backend == CAL.CaltechHPC
     slurm_kwargs = CAL.kwargs(time = 5)
-    slurm_eki = CAL.calibrate(backend, experiment_config; slurm_kwargs)
-    @testset "Caltech HPC Calibration" begin 
+    slurm_eki =
+        CAL.calibrate(backend, experiment_config; slurm_kwargs, model_interface)
+    @testset "Caltech HPC Calibration" begin
         minimal_eki_test(slurm_eki)
     end
 
@@ -120,11 +122,7 @@ end
 # Add plots to help debug
 function scatter_plot(eki::EKP.EnsembleKalmanProcess)
     f = CairoMakie.Figure(resolution = (800, 600))
-    ax = CairoMakie.Axis(
-        f[1, 1],
-        ylabel = "Parameter Value",
-        xlabel = "G",
-    )
+    ax = CairoMakie.Axis(f[1, 1], ylabel = "Parameter Value", xlabel = "G")
 
     g = vec.(EKP.get_g(eki; return_array = true))
     phi = map(x -> abs.(x), vec.((EKP.get_ϕ(prior, eki))))
@@ -142,11 +140,7 @@ scatter_plot(eki)
 
 function phi_versus_iter_plot(eki)
     f = CairoMakie.Figure(resolution = (800, 600))
-    ax = CairoMakie.Axis(
-        f[1, 1],
-        ylabel = "Phi",
-        xlabel = "Iteration",
-    )
+    ax = CairoMakie.Axis(f[1, 1], ylabel = "Phi", xlabel = "Iteration")
     phi = EKP.get_ϕ(prior, eki)
     for (i, phi_) in enumerate(phi)
         CairoMakie.scatter!(ax, fill(i, length(phi_)), vec(phi_))
@@ -160,4 +154,3 @@ function phi_versus_iter_plot(eki)
 end
 
 phi_versus_iter_plot(eki)
-