Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Update to new MOI nonlinear interface #1052

Merged
merged 5 commits into from
Dec 12, 2023
Merged

Update to new MOI nonlinear interface #1052

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
d4addb3
Update to new MOI nonlinear interface
blegat Oct 8, 2023
cea09a9
Fix
blegat Oct 8, 2023
4fa2390
Fixes
blegat Oct 8, 2023
7f59a33
Fixes
blegat Oct 8, 2023
e4e17d3
Fix tests
blegat Oct 9, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
100 changes: 64 additions & 36 deletions src/MOI_wrapper.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,11 +1,10 @@
import MathOptInterface
const MOI = MathOptInterface
import MathOptInterface as MOI

mutable struct Optimizer{T} <: MOI.AbstractOptimizer
# Problem data.
variables::MOI.Utilities.VariablesContainer{T}
starting_values::Vector{Union{Nothing,T}}
nlp_data::Union{MOI.NLPBlockData,Nothing}
nlp_model::Union{MOI.Nonlinear.Model,Nothing}
sense::MOI.OptimizationSense

# Parameters.
Expand Down Expand Up @@ -33,7 +32,9 @@ Optimizer() = Optimizer{Float64}()

MOI.supports(::Optimizer, ::MOI.NLPBlock) = true

MOI.supports(::Optimizer, ::MOI.ObjectiveSense) = true
function MOI.supports(::Optimizer, ::Union{MOI.ObjectiveSense,MOI.ObjectiveFunction})
return true
end
MOI.supports(::Optimizer, ::MOI.Silent) = true
function MOI.supports(::Optimizer, p::MOI.RawOptimizerAttribute)
return p.name == "method" || hasfield(Options, Symbol(p.name))
Expand All @@ -44,6 +45,7 @@ function MOI.supports(::Optimizer, ::MOI.VariablePrimalStart, ::Type{MOI.Variabl
end

const BOUNDS{T} = Union{MOI.LessThan{T},MOI.GreaterThan{T},MOI.EqualTo{T},MOI.Interval{T}}
const _SETS{T} = Union{MOI.GreaterThan{T},MOI.LessThan{T},MOI.EqualTo{T}}

function MOI.supports_constraint(
::Optimizer{T},
Expand All @@ -53,6 +55,14 @@ function MOI.supports_constraint(
return true
end

function MOI.supports_constraint(
::Optimizer{T},
::Type{MOI.ScalarNonlinearFunction},
::Type{<:_SETS{T}},
) where {T}
return true
end

MOI.supports_incremental_interface(::Optimizer) = true

function MOI.copy_to(model::Optimizer, src::MOI.ModelLike)
Expand All @@ -65,6 +75,14 @@ function MOI.set(model::Optimizer, ::MOI.ObjectiveSense, sense::MOI.Optimization
model.sense = sense
return
end
function MOI.set(model::Optimizer, ::MOI.ObjectiveFunction{F}, func::F) where {F}
nl = convert(MOI.ScalarNonlinearFunction, func)
if isnothing(model.nlp_model)
model.nlp_model = MOI.Nonlinear.Model()
end
MOI.Nonlinear.set_objective(model.nlp_model, nl)
return nothing
end

function MOI.set(model::Optimizer, ::MOI.Silent, value::Bool)
model.silent = value
Expand Down Expand Up @@ -112,7 +130,7 @@ MOI.get(model::Optimizer, ::MOI.SolveTimeSec) = time_run(model.results)
function MOI.empty!(model::Optimizer)
MOI.empty!(model.variables)
empty!(model.starting_values)
model.nlp_data = nothing
model.nlp_model = nothing
model.sense = MOI.FEASIBILITY_SENSE
model.results = nothing
return
Expand All @@ -121,12 +139,12 @@ end
function MOI.is_empty(model::Optimizer)
return MOI.is_empty(model.variables) &&
isempty(model.starting_values) &&
model.nlp_data === nothing &&
isnothing(model.nlp_model) &&
model.sense == MOI.FEASIBILITY_SENSE
end

function MOI.add_variable(model::Optimizer{T}) where {T}
push!(model.starting_values, NaN)
push!(model.starting_values, nothing)
return MOI.add_variable(model.variables)
end
function MOI.is_valid(model::Optimizer, index::Union{MOI.VariableIndex,MOI.ConstraintIndex})
Expand All @@ -137,6 +155,18 @@ function MOI.add_constraint(model::Optimizer{T}, vi::MOI.VariableIndex, set::BOU
return MOI.add_constraint(model.variables, vi, set)
end

function MOI.add_constraint(
model::Optimizer{T},
f::MOI.ScalarNonlinearFunction,
s::_SETS{T},
) where {T}
if model.nlp_model === nothing
model.nlp_model = MOI.Nonlinear.Model()
end
index = MOI.Nonlinear.add_constraint(model.nlp_model, f, s)
return MOI.ConstraintIndex{typeof(f),typeof(s)}(index.value)
end

function starting_value(optimizer::Optimizer{T}, i) where {T}
if optimizer.starting_values[i] !== nothing
return optimizer.starting_values[i]
Expand All @@ -157,11 +187,6 @@ function MOI.set(
return
end

function MOI.set(model::Optimizer, ::MOI.NLPBlock, nlp_data::MOI.NLPBlockData)
model.nlp_data = nlp_data
return
end

function requested_features(::ZerothOrderOptimizer, has_constraints)
return Symbol[]
end
Expand Down Expand Up @@ -198,40 +223,47 @@ function sym_sparse_to_dense!(A, I::Vector, nzval)
end

function MOI.optimize!(model::Optimizer{T}) where {T}
num_variables = length(model.starting_values)
backend = MOI.Nonlinear.SparseReverseMode()
vars = MOI.get(model.variables, MOI.ListOfVariableIndices())
evaluator = MOI.Nonlinear.Evaluator(model.nlp_model, backend, vars)
nlp_data = MOI.NLPBlockData(evaluator)

# load parameters
if model.nlp_data === nothing || !model.nlp_data.has_objective
error("An objective should be provided to Optim with `@NLobjective`.")
if isnothing(model.nlp_model)
error("An objective should be provided to Optim with `@objective`.")
end
objective_scale = model.sense == MOI.MAX_SENSE ? -one(T) : one(T)
zero_μ = zeros(T, length(model.nlp_data.constraint_bounds))
zero_μ = zeros(T, length(nlp_data.constraint_bounds))
function f(x)
return objective_scale * MOI.eval_objective(model.nlp_data.evaluator, x)
return objective_scale * MOI.eval_objective(evaluator, x)
end
function g!(G, x)
fill!(G, zero(T))
MOI.eval_objective_gradient(model.nlp_data.evaluator, G, x)
MOI.eval_objective_gradient(evaluator, G, x)
if model.sense == MOI.MAX_SENSE
rmul!(G, objective_scale)
end
return G
end
function h!(H, x)
fill!(H, zero(T))
MOI.eval_hessian_lagrangian(model.nlp_data.evaluator, H_nzval, x, objective_scale, zero_μ)
MOI.eval_hessian_lagrangian(evaluator, H_nzval, x, objective_scale, zero_μ)
sym_sparse_to_dense!(H, hessian_structure, H_nzval)
return H
end

method = model.method
nl_constrained = !isempty(model.nlp_data.constraint_bounds)
features = MOI.features_available(model.nlp_data.evaluator)
nl_constrained = !isempty(nlp_data.constraint_bounds)
features = MOI.features_available(evaluator)
has_bounds = any(vi -> isfinite(model.variables.lower[vi.value]) || isfinite(model.variables.upper[vi.value]), vars)
if method === nothing
if nl_constrained
method = IPNewton()
elseif :Grad in features
if :Hess in features
# FIXME `fallback_method(f, g!, h!)` returns `Newton` but if there
# are variable bounds, `Newton` is not supported. On the other hand,
# `fallback_method(f, g!)` returns `LBFGS` which is supported if `has_bounds`.
if :Hess in features && !has_bounds
method = fallback_method(f, g!, h!)
else
method = fallback_method(f, g!)
Expand All @@ -241,20 +273,15 @@ function MOI.optimize!(model::Optimizer{T}) where {T}
end
end
used_features = requested_features(method, nl_constrained)
MOI.initialize(model.nlp_data.evaluator, used_features)
MOI.initialize(evaluator, used_features)

if :Hess in used_features
hessian_structure = MOI.hessian_lagrangian_structure(model.nlp_data.evaluator)
hessian_structure = MOI.hessian_lagrangian_structure(evaluator)
H_nzval = zeros(T, length(hessian_structure))
end

initial_x = starting_value.(model, eachindex(model.starting_values))
options = copy(model.options)
has_bounds = any(i -> isfinite(model.variables.lower[i]) || isfinite(model.variables.upper[i]), eachindex(model.starting_values))
if has_bounds
lower = [info.lower_bound for info in model.variable_info]
upper = [info.upper_bound for info in model.variable_info]
end
if !nl_constrained && has_bounds && !(method isa IPNewton)
options = Options(; options...)
model.results = optimize(f, g!, model.variables.lower, model.variables.upper, initial_x, Fminbox(method), options; inplace = true)
Expand All @@ -264,26 +291,26 @@ function MOI.optimize!(model::Optimizer{T}) where {T}
options = Options(; options...)
if nl_constrained || has_bounds
if nl_constrained
lc = [b.lower for b in model.nlp_data.constraint_bounds]
uc = [b.upper for b in model.nlp_data.constraint_bounds]
c!(c, x) = MOI.eval_constraint(model.nlp_data.evaluator, c, x)
lc = [b.lower for b in nlp_data.constraint_bounds]
uc = [b.upper for b in nlp_data.constraint_bounds]
c!(c, x) = MOI.eval_constraint(evaluator, c, x)
if !(:Jac in features)
error("Nonlinear constraints should be differentiable to be used with Optim.")
end
if !(:Hess in features)
error("Nonlinear constraints should be twice differentiable to be used with Optim.")
end
jacobian_structure = MOI.jacobian_structure(model.nlp_data.evaluator)
jacobian_structure = MOI.jacobian_structure(evaluator)
J_nzval = zeros(T, length(jacobian_structure))
function jacobian!(J, x)
fill!(J, zero(T))
MOI.eval_constraint_jacobian(model.nlp_data.evaluator, J_nzval, x)
MOI.eval_constraint_jacobian(evaluator, J_nzval, x)
sparse_to_dense!(J, jacobian_structure, J_nzval)
return J
end
function con_hessian!(H, x, λ)
fill!(H, zero(T))
MOI.eval_hessian_lagrangian(model.nlp_data.evaluator, H_nzval, x, zero(T), λ)
MOI.eval_hessian_lagrangian(evaluator, H_nzval, x, zero(T), λ)
sym_sparse_to_dense!(H, hessian_structure, H_nzval)
return H
end
Expand All @@ -293,7 +320,8 @@ function MOI.optimize!(model::Optimizer{T}) where {T}
)
else
@assert has_bounds
c = TwiceDifferentiableConstraints(model.variables.lower, model.variables.upper)
c = TwiceDifferentiableConstraints(
model.variables.lower, model.variables.upper)
end
model.results = optimize(d, c, initial_x, method, options)
else
Expand Down
28 changes: 15 additions & 13 deletions test/MOI_wrapper.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -44,27 +44,29 @@ function test_MOI_Test()
MOI.ObjectiveBound,
MOI.DualObjectiveValue,
MOI.SolverVersion,
MOI.ConstraintDual,
],
),
exclude = String[
# No objective
"test_attribute_SolveTimeSec",
"test_attribute_RawStatusString",
"test_nonlinear_without_objective",
# FIXME INVALID_MODEL should be returned
"test_nonlinear_invalid",
# FIXME The hessian callback for constraints is called with
# `λ = [-Inf, 0.0]` and then we get `NaN`, ...
"hs071",
# There are nonlinear constraints so we need `IPNewton` but `IPNewton` needs a hessian.
"test_nonlinear_hs071_no_hessian",
# FIXME Here there is no hessian but there is a hessian-vector product, can `IPNewton` work with that ?
"test_nonlinear_hs071_hessian_vector_product",
# FIXME needs https://github.com/jump-dev/MathOptInterface.jl/pull/1625
"test_nonlinear_hs071_NLPBlockDual",
# - CachingOptimizer does not throw if optimizer not attached
"test_model_copy_to_UnsupportedAttribute",
"test_model_copy_to_UnsupportedConstraint",
"expression_hs071",
# Terminates with `OTHER_ERROR`
"test_objective_ObjectiveFunction_duplicate_terms",
"test_objective_ObjectiveFunction_constant",
"test_objective_ObjectiveFunction_VariableIndex",
"test_objective_FEASIBILITY_SENSE_clears_objective",
"test_nonlinear_expression_hs109",
"test_objective_qp_ObjectiveFunction_zero_ofdiag",
"test_objective_qp_ObjectiveFunction_edge_cases",
"test_solve_TerminationStatus_DUAL_INFEASIBLE",
"test_solve_result_index",
"test_modification_transform_singlevariable_lessthan",
"test_modification_delete_variables_in_a_batch",
"test_modification_delete_variable_with_single_variable_obj",
],
)
return
Expand Down