Add separate_kwargs_with_semicolon = true to .JuliaFormatter.toml (#764)

.JuliaFormatter.toml

@@ -5,4 +5,5 @@ always_for_in = true
 always_use_return = true
 margin = 80
 remove_extra_newlines = true
+separate_kwargs_with_semicolon = true
 short_to_long_function_def = true

benchmarks/benchmark.jl

@@ -106,13 +106,13 @@ function report(io::IO, filename_A::String, filename_B::String)
     return
 end
 
-filename_A = benchmark(
+filename_A = benchmark(;
     time_limit = 60,
     stopping_rules = [SDDP.BoundStalling(10, 1e-6)],
     duality_handler = SDDP.ContinuousConicDuality(),
 )
 
-filename_B = benchmark(
+filename_B = benchmark(;
     time_limit = 60,
     stopping_rules = [SDDP.BoundStalling(10, 1e-6)],
     duality_handler = SDDP.LagrangianDuality(),

docs/make.jl

@@ -125,10 +125,10 @@ open(joinpath(@__DIR__, "src", "changelog.md"), "r") do in_io
     end
 end
 
-Documenter.makedocs(
+Documenter.makedocs(;
     sitename = "SDDP.jl",
     authors = "Oscar Dowson",
-    format = Documenter.HTML(
+    format = Documenter.HTML(;
         analytics = "G-HZQQDVMPZW",
         # See  https://github.com/JuliaDocs/Documenter.jl/issues/868
         prettyurls = get(ENV, "CI", nothing) == "true",
@@ -180,4 +180,7 @@ Documenter.makedocs(
     doctestfilters = [r"[\s\-]?\d\.\d{6}e[\+\-]\d{2}"],
 )
 
-Documenter.deploydocs(repo = "github.com/odow/SDDP.jl.git", push_preview = true)
+Documenter.deploydocs(;
+    repo = "github.com/odow/SDDP.jl.git",
+    push_preview = true,
+)

docs/src/examples/FAST_production_management.jl

@@ -38,5 +38,5 @@ function fast_production_management(; cut_type)
     @test SDDP.calculate_bound(model) ≈ -23.96 atol = 1e-2
 end
 
-fast_production_management(cut_type = SDDP.SINGLE_CUT)
-fast_production_management(cut_type = SDDP.MULTI_CUT)
+fast_production_management(; cut_type = SDDP.SINGLE_CUT)
+fast_production_management(; cut_type = SDDP.MULTI_CUT)

docs/src/examples/FAST_quickstart.jl

@@ -11,7 +11,7 @@ using SDDP, HiGHS, Test
 
 function fast_quickstart()
     model = SDDP.PolicyGraph(
-        SDDP.LinearGraph(2),
+        SDDP.LinearGraph(2);
         lower_bound = -5,
         optimizer = HiGHS.Optimizer,
     ) do sp, t

docs/src/examples/Hydro_thermal.jl

@@ -66,7 +66,7 @@ graph = SDDP.UnicyclicGraph(0.95; num_nodes = 3)
 # probability mass vector can also be provided.
 
 model = SDDP.PolicyGraph(
-    graph,
+    graph;
     sense = :Min,
     lower_bound = 0.0,
     optimizer = HiGHS.Optimizer,
@@ -90,14 +90,14 @@ end
 # achieved using [`SDDP.train`](@ref). There are many options that can be passed, but
 # `iteration_limit` terminates the training after the prescribed number of SDDP iterations.
 
-SDDP.train(model, iteration_limit = 100)
+SDDP.train(model; iteration_limit = 100)
 
 # ## Simulating the policy
 
 # After training, we can simulate the policy using [`SDDP.simulate`](@ref).
 
 sims = SDDP.simulate(model, 100, [:g_t])
-mu = round(mean([s[1][:g_t] for s in sims]), digits = 2)
+mu = round(mean([s[1][:g_t] for s in sims]); digits = 2)
 println("On average, $(mu) units of thermal are used in the first stage.")
 
 # ## Extracting the water values
@@ -108,4 +108,4 @@ println("On average, $(mu) units of thermal are used in the first stage.")
 # decrease in the expected long-run cost.
 
 V = SDDP.ValueFunction(model[1])
-cost, price = SDDP.evaluate(V, x = 10)
+cost, price = SDDP.evaluate(V; x = 10)

docs/src/examples/StochDynamicProgramming.jl_multistock.jl

@@ -10,7 +10,7 @@
 using SDDP, HiGHS, Test
 
 function test_multistock_example()
-    model = SDDP.LinearPolicyGraph(
+    model = SDDP.LinearPolicyGraph(;
         stages = 5,
         lower_bound = -5.0,
         optimizer = HiGHS.Optimizer,

docs/src/examples/StochDynamicProgramming.jl_stock.jl

@@ -11,7 +11,7 @@ using SDDP, HiGHS, Test
 
 function stock_example()
     model = SDDP.PolicyGraph(
-        SDDP.LinearGraph(5),
+        SDDP.LinearGraph(5);
         lower_bound = -2,
         optimizer = HiGHS.Optimizer,
     ) do sp, stage

docs/src/examples/StructDualDynProg.jl_prob5.2_2stages.jl

@@ -10,7 +10,7 @@
 using SDDP, HiGHS, Test
 
 function test_prob52_2stages()
-    model = SDDP.LinearPolicyGraph(
+    model = SDDP.LinearPolicyGraph(;
         stages = 2,
         lower_bound = 0.0,
         optimizer = HiGHS.Optimizer,

docs/src/examples/StructDualDynProg.jl_prob5.2_3stages.jl

@@ -10,7 +10,7 @@
 using SDDP, HiGHS, Test
 
 function test_prob52_3stages()
-    model = SDDP.LinearPolicyGraph(
+    model = SDDP.LinearPolicyGraph(;
         stages = 3,
         lower_bound = 0.0,
         optimizer = HiGHS.Optimizer,

docs/src/examples/agriculture_mccardle_farm.jl

@@ -56,7 +56,7 @@ function test_mccardle_farm_model()
     ])
 
     model = SDDP.PolicyGraph(
-        graph,
+        graph;
         lower_bound = 0.0,
         optimizer = HiGHS.Optimizer,
     ) do subproblem, index

docs/src/examples/air_conditioning.jl

@@ -21,7 +21,7 @@
 using SDDP, HiGHS, Test
 
 function air_conditioning_model(duality_handler)
-    model = SDDP.LinearPolicyGraph(
+    model = SDDP.LinearPolicyGraph(;
         stages = 3,
         lower_bound = 0.0,
         optimizer = HiGHS.Optimizer,

docs/src/examples/air_conditioning_forward.jl

@@ -10,7 +10,7 @@ import HiGHS
 import Test
 
 function create_air_conditioning_model(; convex::Bool)
-    return SDDP.LinearPolicyGraph(
+    return SDDP.LinearPolicyGraph(;
         stages = 3,
         lower_bound = 0.0,
         optimizer = HiGHS.Optimizer,

docs/src/examples/all_blacks.jl

@@ -11,7 +11,7 @@ function all_blacks()
     ## Number of time periods, number of seats, R_ij = revenue from selling seat
     ## i at time j, offer_ij = whether an offer for seat i will come at time j
     (T, N, R, offer) = (3, 2, [3 3 6; 3 3 6], [1 1 0; 1 0 1])
-    model = SDDP.LinearPolicyGraph(
+    model = SDDP.LinearPolicyGraph(;
         stages = T,
         sense = :Max,
         upper_bound = 100.0,

docs/src/examples/asset_management_simple.jl

@@ -21,7 +21,7 @@ function asset_management_simple()
                 [0.5 0.5; 0.5 0.5],
                 [0.5 0.5; 0.5 0.5],
             ],
-        ),
+        );
         lower_bound = -1_000.0,
         optimizer = HiGHS.Optimizer,
     ) do subproblem, index

docs/src/examples/asset_management_stagewise.jl

@@ -18,7 +18,7 @@ function asset_management_stagewise(; cut_type)
     Phi = [-1, 5]
     Psi = [0.02, 0.0]
 
-    model = SDDP.MarkovianPolicyGraph(
+    model = SDDP.MarkovianPolicyGraph(;
         sense = :Max,
         transition_matrices = Array{Float64,2}[
             [1.0]',
@@ -63,14 +63,14 @@ function asset_management_stagewise(; cut_type)
             if node[1] != 3
                 SDDP.Expectation()
             else
-                SDDP.EAVaR(lambda = 0.5, beta = 0.5)
+                SDDP.EAVaR(; lambda = 0.5, beta = 0.5)
             end
         end,
     )
     @test SDDP.calculate_bound(model) ≈ 1.278 atol = 1e-3
     return
 end
 
-asset_management_stagewise(cut_type = SDDP.SINGLE_CUT)
+asset_management_stagewise(; cut_type = SDDP.SINGLE_CUT)
 
-asset_management_stagewise(cut_type = SDDP.MULTI_CUT)
+asset_management_stagewise(; cut_type = SDDP.MULTI_CUT)

docs/src/examples/belief.jl

@@ -28,7 +28,7 @@ function inventory_management_problem()
     SDDP.add_ambiguity_set(graph, [:Ah, :Bh], 1e2)
 
     model = SDDP.PolicyGraph(
-        graph,
+        graph;
         lower_bound = 0.0,
         optimizer = HiGHS.Optimizer,
     ) do subproblem, node

docs/src/examples/biobjective_hydro.jl

@@ -8,7 +8,7 @@
 using SDDP, HiGHS, Statistics, Test
 
 function biobjective_example()
-    model = SDDP.LinearPolicyGraph(
+    model = SDDP.LinearPolicyGraph(;
         stages = 3,
         lower_bound = 0.0,
         optimizer = HiGHS.Optimizer,
@@ -45,7 +45,7 @@ function biobjective_example()
         end
     end
     pareto_weights =
-        SDDP.train_biobjective(model, solution_limit = 10, iteration_limit = 10)
+        SDDP.train_biobjective(model; solution_limit = 10, iteration_limit = 10)
     solutions = [(k, v) for (k, v) in pareto_weights]
     sort!(solutions; by = x -> x[1])
     @test length(solutions) == 10

docs/src/examples/booking_management.jl

@@ -31,7 +31,7 @@ function booking_management_model(num_days, num_rooms, num_requests)
         end
     end
 
-    return model = SDDP.LinearPolicyGraph(
+    return model = SDDP.LinearPolicyGraph(;
         stages = num_requests,
         upper_bound = max_revenue,
         sense = :Max,

docs/src/examples/generation_expansion.jl

@@ -26,7 +26,7 @@ function generation_expansion(duality_handler)
     penalty = 5e5
     ## Discounting rate
     rho = 0.99
-    model = SDDP.LinearPolicyGraph(
+    model = SDDP.LinearPolicyGraph(;
         stages = 5,
         lower_bound = 0.0,
         optimizer = HiGHS.Optimizer,

docs/src/examples/hydro_valley.jl

@@ -90,7 +90,7 @@ function hydro_valley_model(;
     N = length(valley_chain)
 
     ## Initialise SDDP Model
-    return m = SDDP.MarkovianPolicyGraph(
+    return m = SDDP.MarkovianPolicyGraph(;
         sense = sense,
         lower_bound = lower,
         upper_bound = upper,
@@ -215,47 +215,49 @@ function test_hydro_valley_model()
     Random.seed!(11111)
 
     ## deterministic
-    deterministic_model =
-        hydro_valley_model(hasmarkovprice = false, hasstagewiseinflows = false)
+    deterministic_model = hydro_valley_model(;
+        hasmarkovprice = false,
+        hasstagewiseinflows = false,
+    )
     SDDP.train(
-        deterministic_model,
+        deterministic_model;
         iteration_limit = 10,
         cut_deletion_minimum = 1,
         print_level = 0,
     )
     @test SDDP.calculate_bound(deterministic_model) ≈ 835.0 atol = 1e-3
 
     ## stagewise inflows
-    stagewise_model = hydro_valley_model(hasmarkovprice = false)
-    SDDP.train(stagewise_model, iteration_limit = 20, print_level = 0)
+    stagewise_model = hydro_valley_model(; hasmarkovprice = false)
+    SDDP.train(stagewise_model; iteration_limit = 20, print_level = 0)
     @test SDDP.calculate_bound(stagewise_model) ≈ 838.33 atol = 1e-2
 
     ## Markov prices
-    markov_model = hydro_valley_model(hasstagewiseinflows = false)
-    SDDP.train(markov_model, iteration_limit = 10, print_level = 0)
+    markov_model = hydro_valley_model(; hasstagewiseinflows = false)
+    SDDP.train(markov_model; iteration_limit = 10, print_level = 0)
     @test SDDP.calculate_bound(markov_model) ≈ 851.8 atol = 1e-2
 
     ## stagewise inflows and Markov prices
     markov_stagewise_model =
-        hydro_valley_model(hasstagewiseinflows = true, hasmarkovprice = true)
-    SDDP.train(markov_stagewise_model, iteration_limit = 10, print_level = 0)
+        hydro_valley_model(; hasstagewiseinflows = true, hasmarkovprice = true)
+    SDDP.train(markov_stagewise_model; iteration_limit = 10, print_level = 0)
     @test SDDP.calculate_bound(markov_stagewise_model) ≈ 855.0 atol = 1.0
 
     ## risk averse stagewise inflows and Markov prices
     riskaverse_model = hydro_valley_model()
     SDDP.train(
-        riskaverse_model,
-        risk_measure = SDDP.EAVaR(lambda = 0.5, beta = 0.66),
+        riskaverse_model;
+        risk_measure = SDDP.EAVaR(; lambda = 0.5, beta = 0.66),
         iteration_limit = 10,
         print_level = 0,
     )
     @test SDDP.calculate_bound(riskaverse_model) ≈ 828.157 atol = 1.0
 
     ## stagewise inflows and Markov prices
-    worst_case_model = hydro_valley_model(sense = :Min)
+    worst_case_model = hydro_valley_model(; sense = :Min)
     SDDP.train(
-        worst_case_model,
-        risk_measure = SDDP.EAVaR(lambda = 0.5, beta = 0.0),
+        worst_case_model;
+        risk_measure = SDDP.EAVaR(; lambda = 0.5, beta = 0.0),
         iteration_limit = 10,
         print_level = 0,
     )
@@ -264,26 +266,26 @@ function test_hydro_valley_model()
     ## stagewise inflows and Markov prices
     cutselection_model = hydro_valley_model()
     SDDP.train(
-        cutselection_model,
+        cutselection_model;
         iteration_limit = 10,
         print_level = 0,
         cut_deletion_minimum = 2,
     )
     @test SDDP.calculate_bound(cutselection_model) ≈ 855.0 atol = 1.0
 
     ## Distributionally robust Optimization
-    dro_model = hydro_valley_model(hasmarkovprice = false)
+    dro_model = hydro_valley_model(; hasmarkovprice = false)
     SDDP.train(
-        dro_model,
+        dro_model;
         risk_measure = SDDP.ModifiedChiSquared(sqrt(2 / 3) - 1e-6),
         iteration_limit = 10,
         print_level = 0,
     )
     @test SDDP.calculate_bound(dro_model) ≈ 835.0 atol = 1.0
 
-    dro_model = hydro_valley_model(hasmarkovprice = false)
+    dro_model = hydro_valley_model(; hasmarkovprice = false)
     SDDP.train(
-        dro_model,
+        dro_model;
         risk_measure = SDDP.ModifiedChiSquared(1 / 6),
         iteration_limit = 20,
         print_level = 0,

docs/src/examples/infinite_horizon_hydro_thermal.jl

@@ -53,7 +53,7 @@ function infinite_hydro_thermal(; cut_type)
         model;
         cut_type = cut_type,
         log_frequency = 100,
-        sampling_scheme = SDDP.InSampleMonteCarlo(terminate_on_cycle = true),
+        sampling_scheme = SDDP.InSampleMonteCarlo(; terminate_on_cycle = true),
         cycle_discretization_delta = 0.1,
     )
     @test SDDP.calculate_bound(model) ≈ 119.167 atol = 0.1
@@ -68,5 +68,5 @@ function infinite_hydro_thermal(; cut_type)
     return
 end
 
-infinite_hydro_thermal(cut_type = SDDP.SINGLE_CUT)
-infinite_hydro_thermal(cut_type = SDDP.MULTI_CUT)
+infinite_hydro_thermal(; cut_type = SDDP.SINGLE_CUT)
+infinite_hydro_thermal(; cut_type = SDDP.MULTI_CUT)

docs/src/examples/infinite_horizon_trivial.jl

@@ -14,7 +14,7 @@ function infinite_trivial()
         [(:root_node => :week, 1.0), (:week => :week, 0.9)],
     )
     model = SDDP.PolicyGraph(
-        graph,
+        graph;
         lower_bound = 0.0,
         optimizer = HiGHS.Optimizer,
     ) do subproblem, node

docs/src/examples/no_strong_duality.jl

@@ -16,7 +16,7 @@ function no_strong_duality()
             :root,
             [:node],
             [(:root => :node, 1.0), (:node => :node, 0.5)],
-        ),
+        );
         optimizer = HiGHS.Optimizer,
         lower_bound = 0.0,
     ) do sp, t