added an example

Author: kibaekkim

examples/dcap.jl

@@ -0,0 +1,47 @@
+using StructJuMP
+using Random
+
+function DCAP(nR::Int, nN::Int, nT::Int, nS::Int, seed::Int=1)::StructuredModel
+
+    # set random seed (default=1)
+    Random.seed!(seed)
+
+    # generate & store instance data
+    ## sets
+    R = 1:nR
+    N = 1:nN
+    T = 1:nT
+    S = 1:nS
+
+    ## parameters
+    a = rand(nR, nT) * 5 .+ 5
+    b = rand(nR, nT) * 40 .+ 10
+    c = rand(nR, nN, nT, nS) * 5 .+ 5
+    c0 = rand(nN, nT, nS) * 500 .+ 500
+    d = rand(nN, nT, nS) .+ 0.5
+    Pr = ones(nS)/nS
+
+    # construct JuMP.Model
+    model = StructuredModel(num_scenarios = nS)
+
+    ## 1st stage
+    @variable(model, x[i=R,t=T] >= 0)
+    @variable(model, u[i=R,t=T], Bin)
+    @objective(model, Min, sum(a[i,t]*x[i,t] + b[i,t]*u[i,t] for i in R for t in T))
+    @constraint(model, [i=R,t=T], x[i,t] - u[i,t] <= 0)
+
+    ## 2nd stage
+    for s in S
+        sb = StructuredModel(parent=model, id = s, prob = Pr[s])
+        # @variable(sb, y[i=R, j=N, t=T], Bin)
+        @variable(sb, 0 <= y[i=R, j=N, t=T] <= 1)
+        #@variable(sb, z[j=N,t=T] >= 0) # originally implemented variable (continuous)
+        # @variable(sb, z[j=N,t=T], Bin)  # modify as SIPLIB 1.0
+        @variable(sb, 0 <= z[j=N,t=T] <= 1)
+        @objective(sb, Min, sum(c[i,j,t,s]*y[i,j,t] for i in R for j in N for t in T) + sum(c0[j,t,s]*z[j,t] for j in N for t in T))
+        @constraint(sb, [i=R, t=T], -sum(x[i,tau] for tau in 1:t) + sum(d[j,t,s]*y[i,j,t] for j in N) <= 0)
+        @constraint(sb, [j=N, t=T], sum(y[i,j,t] for i in R) + z[j,t] == 1)
+    end
+
+    return model
+end