Add raw nlp example (#204)

* add raw nlp example * fix test, improve names, generalize deps * Update ci.yml * fix param name * remove log * test with more params * remove param * limit threads * fix type in addmipsol * cleanup
jump-dev · Apr 11, 2023 · 4d70faf · 4d70faf
1 parent bee4112
commit 4d70faf
Show file tree

Hide file tree

Showing 6 changed files with 111 additions and 13 deletions.
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
@@ -32,9 +32,9 @@ jobs:
             ${{ runner.os }}-
       - uses: julia-actions/julia-buildpkg@v1
         env:
-          SECRET_XPRS_WIN_8110: ${{ secrets.XPRS_WIN_8110 }}
-          SECRET_XPRL_WIN_8110: ${{ secrets.XPRL_WIN_8110 }}
-          SECRET_XPRA_WIN_8130: ${{ secrets.XPAUTH_XPR }}
+          SECRET_XPRS_WIN: ${{ secrets.XPRS_WIN_900 }}
+          SECRET_XPRL_WIN: ${{ secrets.XPRL_WIN_900 }}
+          SECRET_XPRA_WIN: ${{ secrets.XPAUTH_XPR }}
       - uses: julia-actions/julia-runtest@v1
       - uses: julia-actions/julia-processcoverage@v1
       - uses: codecov/codecov-action@v1

diff --git a/deps/build.jl b/deps/build.jl
@@ -66,9 +66,9 @@ end
 function ci_installation()
     files = if Sys.iswindows()
     [
-        (ENV["SECRET_XPRS_WIN_8110"], "xprs.dll")
-        (ENV["SECRET_XPRL_WIN_8110"], "xprl.dll")
-        (ENV["SECRET_XPRA_WIN_8130"], "xpauth.xpr")
+        (ENV["SECRET_XPRS_WIN"], "xprs.dll")
+        (ENV["SECRET_XPRL_WIN"], "xprl.dll")
+        (ENV["SECRET_XPRA_WIN"], "xpauth.xpr")
     ]
     end
     for (url, file) in files
@@ -84,12 +84,11 @@ function ci_installation()
     end
 end
 
-
 if haskey(ENV, "XPRESS_JL_SKIP_LIB_CHECK")
     # Skip!
 elseif get(ENV, "JULIA_REGISTRYCI_AUTOMERGE", "false") == "true"
     write_depsfile("julia_registryci_automerge")
-elseif get(ENV, "SECRET_XPRS_WIN_8110", "") != ""
+elseif get(ENV, "SECRET_XPRS_WIN", "") != ""
     ci_installation()
 else
     local_installation()

diff --git a/src/MOI/MOI_callbacks.jl b/src/MOI/MOI_callbacks.jl
@@ -236,7 +236,7 @@ function MOI.submit(
         count += 1
     end
     mipsolcol = Cint.(mipsolcol)
-    mipsolval = Cfloat.(mipsolval)
+    mipsolval = Cdouble.(mipsolval)
     if ilength == MOI.get(model, MOI.NumberOfVariables())
         mipsolcol = C_NULL
     end

diff --git a/src/MOI/MOI_wrapper.jl b/src/MOI/MOI_wrapper.jl
@@ -4275,6 +4275,9 @@ end
 function _pass_variable_names_to_solver(model::Xpress.Optimizer; warn = true)
     NAMELENGTH = 64
     n_variables = length(model.variable_info)
+    if n_variables == 0
+        return nothing
+    end
     var_names = String[string('C', i) for i in 1:n_variables]
     duplicate_check = Set{String}()
     for variable in values(model.variable_info)
@@ -4308,6 +4311,9 @@ end
 function _pass_constraint_names_to_solver(model::Xpress.Optimizer; warn = true)
     NAMELENGTH = 64
     n_constraints = length(model.affine_constraint_info)
+    if n_constraints == 0
+        return nothing
+    end
     con_names = String[string('R', i) for i in 1:n_constraints]
     duplicate_check = Set{String}()
     for constraint in values(model.affine_constraint_info)

diff --git a/test/MathOptInterface/MOI_callbacks.jl b/test/MathOptInterface/MOI_callbacks.jl
@@ -7,8 +7,9 @@ const MOI = MathOptInterface
 
 function callback_simple_model()
     model = Xpress.Optimizer(
-        HEURSTRATEGY = 0,
-        OUTPUTLOG = 0
+        HEURSTRATEGY = 0, # before v41
+        HEUREMPHASIS = 0,
+        OUTPUTLOG = 0,
     )
 
     MOI.Utilities.loadfromstring!(model, """
@@ -26,8 +27,15 @@ end
 
 function callback_knapsack_model()
     model = Xpress.Optimizer(
-        HEURSTRATEGY = 0,
-        OUTPUTLOG = 0
+        OUTPUTLOG    = 0,
+        HEURSTRATEGY = 0, # before v41
+        HEUREMPHASIS = 0,
+        CUTSTRATEGY  = 0,
+        PRESOLVE     = 0,
+        MIPPRESOLVE  = 0,
+        PRESOLVEOPS  = 0,
+        MIPTHREADS = 1,
+        THREADS = 1,
     )
     MOI.set(model, MOI.NumberOfThreads(), 2)
 

diff --git a/test/MathOptInterface/MOI_wrapper.jl b/test/MathOptInterface/MOI_wrapper.jl
@@ -891,6 +891,91 @@ function test_name_empty_names()
     return nothing
 end
 
+function test_dummy_nlp()
+    model = Xpress.Optimizer(OUTPUTLOG = 0);
+    x = MOI.add_variables(model, 2);
+    c = [1.0, 2.0]
+    MOI.set(
+        model,
+        MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}(),
+        MOI.ScalarAffineFunction(MOI.ScalarAffineTerm.(c, x), 0.0),
+    );
+
+    MOI.set(model, MOI.VariableName(), x, ["x1", "x2"])
+
+    Xpress._pass_variable_names_to_solver(model)
+
+    MOI.set(model, MOI.ObjectiveSense(), MOI.MIN_SENSE)
+
+    b1 = MOI.add_constraint.(
+        model,
+        x,
+        MOI.GreaterThan(0.0),
+    )
+    b1 = MOI.add_constraint.(
+        model,
+        x,
+        MOI.LessThan(10.0),
+    )
+
+    c1 = MOI.add_constraint(
+        model,
+        MOI.ScalarAffineFunction(MOI.ScalarAffineTerm.([1.0, 0.0], x), 0.0),
+        MOI.EqualTo(0.0),
+    )
+
+    c3 = MOI.add_constraint(
+        model,
+        MOI.ScalarAffineFunction(MOI.ScalarAffineTerm.([0.0, 1.0], x), 0.0),
+        MOI.GreaterThan(10.0),
+    )
+
+    MOI.optimize!(model)
+
+    x_sol = MOI.get.(model, MOI.VariablePrimal(), x)
+    @test x_sol == [0.0, 10.0]
+
+    ret = Xpress.Lib.XPRSnlpchgformulastring(model.inner, Cint(0), "- 5 * x1 - 3")
+    @test ret == 0
+    ret = Xpress.Lib.XPRSnlpchgformulastring(model.inner, Cint(0), "- 3.14")
+    @test ret == 0
+
+    solvestatus = Ref{Cint}(0)
+    solstatus = Ref{Cint}(0)
+    ret = Xpress.Lib.XPRSoptimize(model.inner, "", solvestatus, solstatus)
+    @test ret == 0
+
+    xx = Array{Float64}(undef, 2)
+    slack = Array{Float64}(undef, 2)
+    duals = Array{Float64}(undef, 2)
+    djs = Array{Float64}(undef, 2)
+    ret = Xpress.Lib.XPRSgetnlpsol(model.inner, xx, slack, duals, djs)
+    @test ret == 0
+    @test xx == [3.14, 10] 
+    @test slack == [0, 0]
+
+    ret = Xpress.Lib.XPRSnlpchgformulastring(model.inner, Cint(0), "- 0.5 * x1 - 3")
+    @test ret == 0
+
+    # to optimize NLPs we need: XPRSoptimize 
+    solvestatus = Ref{Cint}(0)
+    solstatus = Ref{Cint}(0)
+    ret = Xpress.Lib.XPRSoptimize(model.inner, "", solvestatus, solstatus)
+    @test ret == 0
+
+    # to get solution values from NLP we need: XPRSgetnlpsol
+    xx = Array{Float64}(undef, 2)
+    slack = Array{Float64}(undef, 2)
+    duals = Array{Float64}(undef, 2)
+    djs = Array{Float64}(undef, 2)
+    ret = Xpress.Lib.XPRSgetnlpsol(model.inner, xx, slack, duals, djs)
+    @test ret == 0
+    @test xx == [6, 10] 
+    @test slack == [0, 0]
+
+    return nothing
+end
+
 end
 
 TestMOIWrapper.runtests()