[UPD] Replace simulation_scenarios call in src/

Author: frapac

src/SDDPoptimize.jl

@@ -88,9 +88,9 @@ function run_SDDP(model::SPModel,
 
     while (iteration_count < param.maxItNumber) & (~stopping_test)
         # Time execution of current pass:
-		if display > 0
-			tic()
-		end
+        if display > 0
+            tic()
+        end
 
         # Build given number of scenarios according to distribution
         # law specified in model.noises:
@@ -171,10 +171,7 @@ Float64 (estimation of the upper bound)
 """
 function estimate_upper_bound(model, param, V, problems, n_simulation=1000)
 
-    aleas = simulate_scenarios(model.noises ,
-                                    (model.stageNumber-1,
-                                     n_simulation,
-                                     model.dimNoises))
+    aleas = simulate_scenarios(model.noises, n_simulation)
 
     costs, stockTrajectories, _ = forward_simulations(model,
                                                         param,
@@ -332,10 +329,7 @@ function initialize_value_functions( model::SPModel,
     V = Array{PolyhedralFunction}(model.stageNumber)
 
     # Build scenarios according to distribution laws:
-    aleas = simulate_scenarios(model.noises,
-                               (model.stageNumber-1,
-                                param.forwardPassNumber,
-                                model.dimNoises))
+    aleas = simulate_scenarios(model.noises, param.forwardPassNumber)
 
     V[end] = V_final
 
@@ -464,20 +458,20 @@ Parameters:
     Linear problems used to approximate the value functions
 
 - t (Int64)
-	Time
+    Time
 
 - xt (Vector{Float64})
-	Position where to compute optimal control
+    Position where to compute optimal control
 
 - xi (Vector{Float64})
-	Alea at time t
+    Alea at time t
 
 Return:
-	Vector{Float64}: optimal control at time t
+    Vector{Float64}: optimal control at time t
 
 """
 function get_control(model::SPModel, param::SDDPparameters, lpproblem::Vector{JuMP.Model}, t, xt, xi)
-	return solve_one_step_one_alea(model, param, lpproblem[t], t, xt, xi)[2].optimal_control
+    return solve_one_step_one_alea(model, param, lpproblem[t], t, xt, xi)[2].optimal_control
 end
 
 

test/runtests.jl

@@ -22,13 +22,10 @@ facts("Probability functions") do
     @fact law.supportSize --> 3
 
     dims = (2, 2, 1)
-    scenarios = simulate_scenarios([law, law], dims)
+    scenarios = simulate_scenarios([law, law], 2)
     @fact typeof(scenarios) --> Array{Float64, 3}
     @fact size(scenarios) --> (2, 2, 1)
 
-    scenarios2 = simulate_scenarios(Normal(), dims)
-    @fact typeof(scenarios2) --> Array{Float64, 3}
-
     # test product of noiselaws:
     support2 = [4, 5, 6]
     proba2 = [.3 .3 .4]
@@ -109,10 +106,7 @@ facts("SDDP algorithm: 1D case") do
     cost,
     dynamic, laws)
     # Generate scenarios for forward simulations:
-    noise_scenarios = simulate_scenarios(model.noises,
-    (model.stageNumber,
-    params.forwardPassNumber,
-    model.dimNoises))
+    noise_scenarios = simulate_scenarios(model.noises,params.forwardPassNumber)
 
     sddp_costs = 0
     context("Linear cost") do
@@ -250,10 +244,7 @@ facts("SDDP algorithm: 2D case") do
 
 
         # Test a simulation upon given scenarios:
-        noise_scenarios = simulate_scenarios(model.noises,
-        (model.stageNumber,
-        n_simulations,
-        model.dimNoises))
+        noise_scenarios = simulate_scenarios(model.noises,n_simulations)
 
         sddp_costs, stocks = forward_simulations(model, params, V, pbs, noise_scenarios)
 
@@ -445,8 +436,8 @@ facts("SDP algorithm") do
 
             costs_sdp, stocks_sdp, controls_sdp = sdp_forward_simulation(modelSDP,
             paramsSDP,
-            aleas_scen, x0,
-            V_sdp, true )
+            aleas_scen,
+            V_sdp, true)
 
             @fact size(stocks_sdp) --> (3,1,2)
             @fact size(controls_sdp) --> (2,1,2)