v1.29.3

JuMP v1.29.3

Diff since v1.29.2

Merged pull requests:

• [Containers] fix support for sum(::DenseAxisArray; init) (#4085) (@odow)
• [docs] remove DSDP from the documentation (#4086) (@odow)
• [docs] add UnoSolver.jl to the documentation (#4087) (@odow)
• Fix set_start_values with parameters (#4090) (@odow)
• Prep for v1.29.3 (#4091) (@odow)

Closed issues:

• UnderKeywordError when getting constraints from MIPLIB instance (#4084)
• Error when using set_start_values with models containing parameters (#4089)

v1.29.2

JuMP v1.29.2

Diff since v1.29.1

Merged pull requests:

• fix: add necessary promote_operation methods (#4071) (@AayushSabharwal)
• Update parallelism.md (#4072) (@WalterMadelim)
• Update column generation tutorial (#4074) (@sandyspiers)
• [docs] update package versions (#4075) (@odow)
• Add spacing after min in OPF tutorial (#4077) (@blegat)
• Nicer real and imag in LaTeX (#4078) (@blegat)
• Nicer imaginary number in LaTeX (#4079) (@blegat)
• Move to JuliaFormatter@v2 (#4080) (@blegat)
• Prep for v1.29.2 (#4081) (@odow)
• Relax tests for upcoming MOI release (#4082) (@odow)
• [docs] pin to use Julia v1.11 (#4083) (@odow)

Closed issues:

• Why does JuMP maintain DenseAxisArray instead of using AxisKeys.jl for named (incl. integer) indexing? (#4070)
• Adding set_integer(::AffExpr)? (#4073)
• Unclear complementarity in example (#4076)

v1.29.1

JuMP v1.29.1

Diff since v1.29.0

Merged pull requests:

• prevent adding methods to the functions > and >= (#4065) (@nsajko)
• Prep for v1.29.1 (#4069) (@odow)

v1.29.0

JuMP v1.29.0

Diff since v1.28.0

Merged pull requests:

• [docs] remove mentions of Gitter (#4045) (@odow)
• Add simplify, derivative, and gradient from MOI.Nonlinear.SymbolicAD (#4047) (@odow)
• Fix promote_shape for DenseAxisArrray (#4050) (@odow)
• Fix jump_function to promote model and coefficient types (#4051) (@odow)
• [Containers] fix DenseAxisArray with Base.OneTo{<:Integer} (#4054) (@odow)
• [Containers] fix axis lookup for AbstractUnitRange axes (#4055) (@odow)
• [Containers] fix getindex when a custom subtype of <:Integer is used (#4056) (@odow)
• Add missing promote_operation when mixing number types (#4057) (@blegat)
• Disable depwarn in ci (#4058) (@blegat)
• [docs] document closure capture bug in threading (#4059) (@odow)
• Prep for v1.29.0 (#4060) (@odow)
• Add cuOpt to list of solvers (#4062) (@blegat)
• Add quick reference to supported solvers (#4063) (@langestefan)
• [docs] clarify that JuMP is not a CAS (#4064) (@odow)

Closed issues:

• Vectorized set_lower_bound and set_upper_bound (#4030)
• Evaluating nonlinear expressions with variable references (#4044)
• mayra-salgado-hhhhhgghhhh (#4046)
• jump_function ignores coefficient type (#4048)
• Better error than promote_shape ? (#4049)
• Issues using DenseAxisArrays with typed indexing (#4053)

v1.28.0

JuMP v1.28.0

Diff since v1.27.0

Merged pull requests:

• [docs] better method for creating empty array (#4031) (@araujoms)
• Fix a major correctess bug with comparison legacy nonlinear constraints (#4037) (@odow)
• Add support for variables in Nonnegatives,Nonpositives,Zeros (#4038) (@odow)
• [docs] update package versions (#4039) (@odow)
• Don't copy objective in copy_conflict (#4041) (@odow)
• [docs] update docs because HiGHS now supports an IIS (#4042) (@odow)
• Prep for v1.28.0 (#4043) (@odow)

Closed issues:

• Vectorized set_lower_bound and set_upper_bound (#4030)
• Design patterns for larger models: reading using JSON3, separate model building, solving and results and a struct for variables? (#4034)
• Nonnegative variables (#4035)
• Don't add objective in copy_conflict (#4040)

v1.27.0

JuMP v1.27.0

Diff since v1.26.0

Merged pull requests:

• use new syntax in state discrimination tutorial (#4009) (@araujoms)
• add value(::AbstractArray) (#4010) (@araujoms)
• update MATLAB tutorial (#4012) (@araujoms)
• [docs] fix doc linking of DenseAxisArray and SparseAxisArray (#4013) (@odow)
• [docs] add sparsity section to performance tips (#4015) (@odow)
• Improve the type stability of constant powers (#4016) (@odow)
• [docs] add a tutorial on typed indices (#4022) (@odow)
• Fix expression comparisons for Julia nightly (#4023) (@odow)
• [docs] update tolerances.jl tutorial (#4025) (@odow)
• Mention set_objective_sense in documentation manual (#4028) (@blegat)
• Prep for v1.27.0 (#4029) (@odow)

Closed issues:

• SparseArrays.SparseMatrixCSC support (#4005)
• Poor performance with many QuadExpr operations (#4011)
• Controlling Algorithm Selection in Gurobi for MILP Models in Julia's JuMP (#4018)
• Help Needed: Feasibility Cuts Code Error in JuMP's Benders Decomposition Documentation (#4019)
• Typed ints for indices (#4020)
• SCIP does not seem to allow user defined operators (#4026)
• Documentation erratum set new objective (#4027)

v1.26.0

JuMP v1.26.0

Diff since v1.25.0

Merged pull requests:

• Add set_name(model::GenericModel, name::AbstractString) (#3967) (@odow)
• Simplify implementation of shadow_price and improve docstring (#3968) (@odow)
• Improve formatting by adding blank lines between functions (#3969) (@odow)
• Replace diagm by Diagonal (#3971) (@blegat)
• Update to cache v2 (#3972) (@blegat)
• Catch and rethrow more informative error for BoundAlreadySet in optimize! (#3974) (@odow)
• [docs] switch to using secrets.GUROBI_LICENSE (#3975) (@odow)
• [docs] update versions in Project.toml and packages.toml (#3976) (@odow)
• [docs] remove admonitions from callack manual page (#3977) (@odow)
• Add support for with_cache_type in Model and set_optimizer (#3979) (@odow)
• Add print_macro_timing_summary (#3981) (@odow)
• Fix add_to_expression! with GenericAffExpr{ComplexF64} (#3983) (@odow)
• Fix docstring of set_normalized_coefficient (#3985) (@odow)
• Prep for v1.26.0 (#3986) (@odow)
• [docs] add conventions page to the developer docs (#3987) (@odow)
• Fix reshaping ::Nothing for matrix shapes (#3988) (@odow)
• [docs] add developer docs for unset_ methods (#3990) (@odow)
• Do not set variable name when unsupported by solver in _moi_add_constrained_variables (#3995) (@guimarqu)
• Fix show of SparseAxisArray for small display size (#3996) (@blegat)
• Fix LinearAlgebra.issymmetric(::AbstractJuMPScalar) (#3997) (@odow)
• Add value(::Function, ::X) and value(::X; result::Int) for Symmetric and Hermitian (#4000) (@araujoms)
• Update cutting_stock_column_generation.jl (#4001) (@WalterMadelim)
• Fix missing value_type methods for expressions (#4002) (@odow)
• [docs] add example of using SparseArrays to variable manual (#4003) (@odow)
• [docs] add MathOptChordalDecomposition (#4004) (@odow)
• fix value for partially initialized Hermitian matrices (#4007) (@araujoms)
• [docs] use GITHUB_TOKEN when downloading from GitHub (#4008) (@odow)

Closed issues:

• Error when adding a Nonpositive constraint in a variable bounded above (#3973)
• cache option for set_optimizer (#3978)
• Constraint attribute are not converted with moi_function (#3980)
• Can't multiply a Matrix{AffExpr} by a Matrix{ComplexF64} (#3982)
• set_normalized_coefficient fails with vector inputs (#3984)
• Add dev docs for functions that activate options in JuMP (#3989)
• Unexpected behavior when adding constrained variable to a solver that does not support name (direct model) (#3994)
• value doesn't work for complex matrices (#3998)
• Add more methods for value_type(::Type{<:AbstractJuMPScalar}) (#3999)
• value crashes on partially initialized matrices (#4006)

v1.25.0

JuMP v1.25.0

Diff since v1.24.0

Merged pull requests:

• [docs] add a section on Function barriers to design patterns tutorial (#3938) (@odow)
• [docs] clarify how to pass PSD and Symmetric via set kwarg (#3940) (@odow)
• [docs] use julia-actions/cache (#3942) (@odow)
• [docs] Fix typo in psd docs (#3943) (@joaquimg)
• [docs] change to use Clarabel in more tutorials (#3944) (@odow)
• [docs] add section on unbounded rays (#3945) (@odow)
• [docs] update version in docs/packages.toml (#3949) (@odow)
• [docs] expand section on the limitations of parameters (#3950) (@odow)
• Remove unreachable method of add_variable (#3952) (@odow)
• Improve test coverage (#3953) (@odow)
• Modify usage of PrecompileTools.jl to improve code coverage (#3954) (@odow)
• Add VariableInSetRef and is_variable_in_set (#3955) (@odow)
• [docs] improve docstrings of write_to_file and read_from_file (#3956) (@odow)
• [docs] fix: Add correct parameters for set_attribute in installation.md (#3957) (@Prasanthi-Peram)
• [docs] mention use_nlp_block when reading nonlinear file formats (#3958) (@odow)
• [docs] use VariableInSetRef in the tutorials (#3960) (@odow)
• [docs] remove has_values and has_duals from documentation (#3961) (@odow)
• Prep for v1.25.0 (#3962) (@odow)
• Refactor show(::IO, ::_SolutionSummary) into a tree structure (#3963) (@odow)
• [docs] update dead gurobi link (#3964) (@ArnoStrouwen)
• [docs] bump versions in Project.toml and docs/packages.toml (#3965) (@odow)
• [docs] clarify behavior of FEASIBILITY_SENSE in documentation (#3966) (@odow)

Closed issues:

• Macro to update expression in-place (#3931)
• Allocations in MA (or possibly JuMP) (#3936)
• Performance tip for model's getindex (#3937)
• Model gets stuck building a nested user-defined function (#3939)
• Unexpected interactions with JuMP.Parameter (in QPs) (#3947)
• Add option to print the solvers in memory representation (#3948)
• Constraint associated with a variable set (#3951)

v1.24.0

JuMP v1.24.0

Diff since v1.23.6

Merged pull requests:

• [docs] add danger admonition for Threads.threadid (#3909) (@odow)
• Remove warning from operator_warn (#3911) (@odow)
• Fix primal_feasibility_report with non-Float64 number types (#3913) (@odow)
• [docs] use tectonic_jll (#3914) (@odow)
• Update variables.md explanation of parameters (#3916) (@artkuo)
• [docs] exclude Gurobi examples if built from fork (#3919) (@odow)
• DNM: test #3919 (#3920) (@odow)
• [docs] fix Gurobi examples when built from fork (#3924) (@odow)
• Add assert_is_solved_and_feasible (#3925) (@odow)
• [docs] fix year in changelog.md (#3926) (@odow)
• [docs] add links in SecondOrderCone docstring (#3927) (@odow)
• [docs] update docs/Project.toml (#3928) (@odow)
• [docs] remove mention of the time-to-first-solve issue (#3929) (@odow)
• [docs] elaborate on add_to_expression in performance tips (#3930) (@odow)
• [docs] update docstring of is_solved_and_feasible (#3932) (@odow)
• Fix bug copying expressions that do not contain variables (#3934) (@odow)
• Prep for v1.24.0 (#3935) (@odow)

Closed issues:

• Nonlinear subexpressions (#3738)
• Remove the warning in operator_warn (#3910)
• Type error for primal feasibility report in non-Float64 precision (#3912)
• Gurobi in the docs means forks will fail (#3917)
• NEWS.md confuses automated tooling (#3918)
• assert_is_solved_and_feasible (#3923)
• Copy : unsupported subtype of DenseAxisArray (#3933)

v1.23.6

JuMP v1.23.6

Diff since v1.23.5

Merged pull requests:

• [docs] add Norm{Nuclear,Spectral}Cone examples (#3887) (@odow)
• Add simple example "Low-rank matrix completion" (#3888) (@jd-foster)
• Fix StackOverflow error in value of GenericNonlinearExpr (#3889) (@odow)
• Improve error message when = is used instead of == (#3892) (@odow)
• Add ellipse fitting example (#3895) (@langestefan)
• [docs] restructure the Parallelism tutorial and add section on model building (#3897) (@odow)
• [docs] fix doctest in @force_nonlinear (#3898) (@odow)
• [docs] call optimize! in solve_and_print_solution (#3899) (@lamont-granquist)
• [docs] add new solver CATrustRegionMethod.jl (#3900) (@fadihamad94)
• [docs] clarify black-box opt. in "Should you use JuMP" (#3901) (@odow)
• [docs] add Gurobi to the docs and remove GLPK (#3904) (@odow)
• [docs] add tutorial on multiple solutions with Gurobi (#3905) (@odow)
• [docs] tweaks to the tolerance tutorial (#3906) (@odow)
• Fix GitHub actions badge in README (#3907) (@odow)
• Prep for v1.23.6 (#3908) (@odow)

Closed issues:

• Stack overflow error when find the value of deeply nested nonlinear expressions (#3884)
• Improve error message when confusing = with == in @constraint? (#3891)
• Ellipse fitting example (#3893)
• Parallel model building docs (#3894)
• JuMP rejects my non-linear objective for Hypatia but Convex.jl doesn't (#3896)
• How to get the extreme rays of a model, please (#3902)
• Log an optimizer's output to somewhere else, instead of simply in Julia REPL? (#3903)