# Updating to MosekToolsv0.15.3 breaks many JuMP SDP codes

**URL:** <https://discourse.julialang.org/t/updating-to-mosektoolsv0-15-3-breaks-many-jump-sdp-codes/123893>\
**Category:** Optimization (Mathematical)\
**Tags:** jump, mosek\
**Created:** [December 16, 2024, 5:04pm UTC](https://discourse.julialang.org/t/updating-to-mosektoolsv0-15-3-breaks-many-jump-sdp-codes/123893 "2024-12-16T17:04:44Z")\
**Posts on this page:** 6\
**Page:** 1

<div class="post-metadata">

**Author:** ![Shuvomoy\_Das\_Gupta](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/shuvomoy_das_gupta/32/10069_2.png) [@Shuvomoy\_Das\_Gupta](https://discourse.julialang.org/u/Shuvomoy_Das_Gupta)\
**Post date:** [December 16, 2024, 5:04pm UTC](https://discourse.julialang.org/t/updating-to-mosektoolsv0-15-3-breaks-many-jump-sdp-codes/123893/1 "2024-12-16T17:04:44Z")

</div>

I updated MosekTools to `v0.15.3` today (from 0.15.1) and it breaks my existing SDP codes with the following error when calling `optimize!`:

```julia
MSK_RES_ERR_SYM_MAT_INVALID_ROW_INDEX): The row index 11 specified for sparse symmetric is invalid. The valid range is [0,10[

```

or another error:

```julia
MOSEK error 3944 (MSK_RES_ERR_SYM_MAT_DUPLICATE): The value (4,4) in a sparse symmetric has been specified twice at position 0 and position 3.

```

(The numbers `11`, `(4,4)`, `0`, `3` etc change from instance to instance)

No idea what is going on, it also breaks so many of my previous SDP codes as well (some of them were written in 2022-23 and ran without any issue).

After downgrading MosekTools to 0.15.1 my previous code works without any problem. I am using macOS and the latest version of JuMP 1.23.5 and Julia (1.11.2+0.aarch64.apple.darwin14) and here are my currently installed packages (only MosekTools is downgraded, rest are updated):

```julia
  [6e4b80f9] BenchmarkTools v1.5.0
  [31c24e10] Distributions v0.25.114
  [2e9cd046] Gurobi v1.6.0
  [b6b21f68] Ipopt v1.7.0
  [033835bb] JLD2 v0.5.10
  [4076af6c] JuMP v1.23.5
  [67920dd8] KNITRO v0.14.4
  [6405355b] Mosek v10.2.0
⌃ [1ec41992] MosekTools v0.15.1
  [6fe1bfb0] OffsetArrays v1.14.2
  [bac558e1] OrderedCollections v1.7.0
  [37e2e46d] LinearAlgebra v1.11.0

```

I was able to reproduce it on a Windows computer as well. Not sure if this is useful: I am using `OffsetArrays` in creating some of the data matrices for the SDPs in consideration and maybe the newly added `VariableBasisStatus` since `MosekTools 0.15.2` is somehow responsible for displaying this error. Any help is much appreciated.

---

<div class="post-metadata">

**Author:** ![odow](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/odow/32/28685_2.png) [@odow](https://discourse.julialang.org/u/odow)\
**Post date:** [December 16, 2024, 9:13pm UTC](https://discourse.julialang.org/t/updating-to-mosektoolsv0-15-3-breaks-many-jump-sdp-codes/123893/2 "2024-12-16T21:13:06Z")

</div>

Please open an issue with a reproducible example. It looks like @blegat made some changes that seem related: [Comparing v0.15.1...v0.15.3 · jump-dev/MosekTools.jl · GitHub](https://github.com/jump-dev/MosekTools.jl/compare/v0.15.1...v0.15.3)

---

<div class="post-metadata">

**Author:** ![Shuvomoy\_Das\_Gupta](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/shuvomoy_das_gupta/32/10069_2.png) [@Shuvomoy\_Das\_Gupta](https://discourse.julialang.org/u/Shuvomoy_Das_Gupta)\
**Post date:** [December 17, 2024, 1:25am UTC](https://discourse.julialang.org/t/updating-to-mosektoolsv0-15-3-breaks-many-jump-sdp-codes/123893/3 "2024-12-17T01:25:00Z")

</div>

Thanks for your response @odow! Here is a reproducible example along with output with both `MosekTools 0.15.3` and `MosekTools 0.15.1`. I have opened an issue on Github as well. Any tips to fix this will be much appreciated (for now I am just using 0.15.1) because this ends up breaking many of my previous SDP codes in existing research projects.

```julia-auto
## Load the packages
# -------------------------

using JuMP, MosekTools, Mosek, LinearAlgebra, OffsetArrays

## Data matrices in OffsetArray format 
# ------------------------------------------------

A = [0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0; 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0; 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.5 0.0 0.0 0.0; 0.0 0.0 0.0 -0.0 0.0 0.0 0.0 0.0 -0.0 -0.5 0.0 -0.0 0.0; 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0; 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0; 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0; 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0; 0.0 0.0 0.0 -0.0 0.0 0.0 0.0 0.0 -0.0 -0.5 0.0 -0.0 0.0; 0.0 0.0 0.5 -0.5 0.0 0.0 0.0 0.0 -0.5 0.0 0.0 -0.5 0.0; 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0; 0.0 0.0 0.0 -0.0 0.0 0.0 0.0 0.0 -0.0 -0.5 0.0 -0.0 0.0; 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0];

A_Offset = OffsetArray(A, 1:13, 1:13)

B = [0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0; 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0; 0.0 0.0 0.0 0.0 0.5 0.0 0.0 0.0; 0.0 0.0 0.0 -0.0 -0.453602203155747 0.0 0.0 0.0; 0.0 0.0 0.5 -0.453602203155747 0.0 0.0 0.0 0.453602; 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0; 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0; 0.0 0.0 0.0 0.0 0.453602 0.0 0.0 0.0];

B_Offset = OffsetArray(B, 1:8, 1:8)

## JuMP model
# -----------------

model_test = Model(Mosek.Optimizer)

@variable(model_test, X_var[1:13, 1:13], PSD)

@variable(model_test, Y_var[1:8, 1:8], PSD)

@objective(model_test, Min, tr(X_var))

@constraint(model_test,tr(X_var * A_Offset) - tr(Y_var * B_Offset) == 0) # This constraint causes the issue

@constraint(model_test, tr(X_var) <= 1)

optimize!(model_test)

```

## OUTPUT with MosekTools v0.15.3:

(Please see output with MosekTools v0.15.1 below, which runs fine)

```julia-auto
MOSEK error 3940 (MSK_RES_ERR_SYM_MAT_INVALID_ROW_INDEX): The row index 9 specified for sparse symmetric is invalid. The valid range is [0,8[.
MOSEK error 3940 (MSK_RES_ERR_SYM_MAT_INVALID_ROW_INDEX): The row index 9 specified for sparse symmetric is invalid. The valid range is [0,8[.
MOSEK error 3940 (MSK_RES_ERR_SYM_MAT_INVALID_ROW_INDEX): The row index 9 specified for sparse symmetric is invalid. The valid range is [0,8[.
MOSEK error 3940 (MSK_RES_ERR_SYM_MAT_INVALID_ROW_INDEX): The row index 11 specified for sparse symmetric is invalid. The valid range is [0,8[.
ERROR: MosekError(3940, "The row index 11 specified for sparse symmetric is invalid. The valid range is [0,8[.\n")
Stacktrace:
  [1] appendsparsesymmat(task::Mosek.Task, dim::Int32, subi::Vector{Int32}, subj::Vector{Int32}, valij::Vector{Float64})
    @ Mosek ~/.julia/packages/Mosek/cMwUF/src/msk_functions.jl:3707
  [2] appendsparsesymmat(task::Mosek.Task, dim::Int64, subi::Vector{Int32}, subj::Vector{Int32}, valij::Vector{Float64})
    @ Mosek ~/.julia/packages/Mosek/cMwUF/src/msk_functions.jl:16628
  [3] (::MosekTools.var"#add_sd#24"{MosekTools.Optimizer, MosekTools.var"#25#26"{…}, Vector{…}, Vector{…}, Vector{…}})()
    @ MosekTools ~/.julia/packages/MosekTools/uICAS/src/constraint.jl:69
  [4] split_scalar_matrix(m::MosekTools.Optimizer, terms::Vector{…}, set_sd::MosekTools.var"#25#26"{…})
    @ MosekTools ~/.julia/packages/MosekTools/uICAS/src/constraint.jl:93
  [5] set_row
    @ ~/.julia/packages/MosekTools/uICAS/src/constraint.jl:103 [inlined]
  [6] add_constraint(m::MosekTools.Optimizer, axb::MathOptInterface.ScalarAffineFunction{…}, dom::MathOptInterface.EqualTo{…})
    @ MosekTools ~/.julia/packages/MosekTools/uICAS/src/constraint.jl:437
  [7] _copy_constraints(dest::MosekTools.Optimizer, src::MathOptInterface.Utilities.UniversalFallback{…}, index_map::MathOptInterface.Utilities.IndexMap, index_map_FS::MathOptInterface.Utilities.DoubleDicts.IndexDoubleDictInner{…}, cis_src::Vector{…})
    @ MathOptInterface.Utilities ~/.julia/packages/MathOptInterface/gLl4d/src/Utilities/copy.jl:180
  [8] _copy_constraints(dest::MosekTools.Optimizer, src::MathOptInterface.Utilities.UniversalFallback{…}, index_map::MathOptInterface.Utilities.IndexMap, cis_src::Vector{…})
    @ MathOptInterface.Utilities ~/.julia/packages/MathOptInterface/gLl4d/src/Utilities/copy.jl:192
  [9] pass_nonvariable_constraints_fallback(dest::MosekTools.Optimizer, src::MathOptInterface.Utilities.UniversalFallback{…}, index_map::MathOptInterface.Utilities.IndexMap, constraint_types::Vector{…})
    @ MathOptInterface.Utilities ~/.julia/packages/MathOptInterface/gLl4d/src/Utilities/copy.jl:203
 [10] pass_nonvariable_constraints
    @ ~/.julia/packages/MathOptInterface/gLl4d/src/Utilities/copy.jl:229 [inlined]
 [11] pass_nonvariable_constraints(dest::MathOptInterface.Bridges.LazyBridgeOptimizer{…}, src::MathOptInterface.Utilities.UniversalFallback{…}, idxmap::MathOptInterface.Utilities.IndexMap, constraint_types::Vector{…})
    @ MathOptInterface.Bridges ~/.julia/packages/MathOptInterface/gLl4d/src/Bridges/bridge_optimizer.jl:426
 [12] _pass_constraints(dest::MathOptInterface.Bridges.LazyBridgeOptimizer{…}, src::MathOptInterface.Utilities.UniversalFallback{…}, index_map::MathOptInterface.Utilities.IndexMap, variable_constraints_not_added::Vector{…})
    @ MathOptInterface.Utilities ~/.julia/packages/MathOptInterface/gLl4d/src/Utilities/copy.jl:251
 [13] default_copy_to(dest::MathOptInterface.Bridges.LazyBridgeOptimizer{…}, src::MathOptInterface.Utilities.UniversalFallback{…})
    @ MathOptInterface.Utilities ~/.julia/packages/MathOptInterface/gLl4d/src/Utilities/copy.jl:393
 [14] copy_to
    @ ~/.julia/packages/MathOptInterface/gLl4d/src/Bridges/bridge_optimizer.jl:442 [inlined]
 [15] optimize!
    @ ~/.julia/packages/MathOptInterface/gLl4d/src/MathOptInterface.jl:121 [inlined]
 [16] optimize!(m::MathOptInterface.Utilities.CachingOptimizer{…})
    @ MathOptInterface.Utilities ~/.julia/packages/MathOptInterface/gLl4d/src/Utilities/cachingoptimizer.jl:321
 [17] optimize!(model::Model; ignore_optimize_hook::Bool, _differentiation_backend::MathOptInterface.Nonlinear.SparseReverseMode, kwargs::@Kwargs{})
    @ JuMP ~/.julia/packages/JuMP/i68GU/src/optimizer_interface.jl:595
 [18] optimize!(model::Model)
    @ JuMP ~/.julia/packages/JuMP/i68GU/src/optimizer_interface.jl:546
 [19] top-level scope
    @ ~/Library/MyDrive/junk_3.jl:35

```

## Output with MosekTools v0.15.1:

```julia-auto
Problem
  Name :                 
  Objective sense : minimize        
  Type : CONIC (conic optimization problem)
  Constraints : 2               
  Affine conic cons. : 0               
  Disjunctive cons. : 0               
  Cones : 0               
  Scalar variables : 0               
  Matrix variables : 2 (scalarized: 127)
  Integer variables : 0               

Optimizer started.
Presolve started.
Linear dependency checker started.
Linear dependency checker terminated.
Eliminator started.
Freed constraints in eliminator : 0
Eliminator terminated.
Eliminator started.
Freed constraints in eliminator : 0
Eliminator terminated.
Eliminator - tries : 2 time : 0.00            
Lin. dep. - tries : 1 time : 0.00            
Lin. dep. - primal attempts : 1 successes : 1               
Lin. dep. - dual attempts : 0 successes : 0               
Lin. dep. - primal deps. : 0 dual deps. : 0               
Presolve terminated. Time: 0.00    
Optimizer - threads : 20              
Optimizer - solved problem : the primal      
Optimizer - Constraints : 2               
Optimizer - Cones : 0               
Optimizer - Scalar variables : 1 conic : 0               
Optimizer - Semi-definite variables: 2 scalarized : 127             
Factor - setup time : 0.00            
Factor - dense det. time : 0.00 GP order time : 0.00            
Factor - nonzeros before factor : 3 after factor : 3               
Factor - dense dim. : 0 flops : 1.20e+03        
ITE PFEAS DFEAS GFEAS PRSTATUS POBJ DOBJ MU TIME  
0 1.2e+01 1.0e+00 1.4e+01 0.00e+00 1.300000000e+01 0.000000000e+00 1.0e+00 0.00  
1 0.0e+00 0.0e+00 0.0e+00 1.00e+00 0.000000000e+00 0.000000000e+00 0.0e+00 0.01  
Optimizer terminated. Time: 0.01

```

---

<div class="post-metadata">

**Author:** ![odow](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/odow/32/28685_2.png) [@odow](https://discourse.julialang.org/u/odow)\
**Post date:** [December 17, 2024, 5:13am UTC](https://discourse.julialang.org/t/updating-to-mosektoolsv0-15-3-breaks-many-jump-sdp-codes/123893/4 "2024-12-17T05:13:32Z")

</div>

There’s no work-around. This just looks like a bug in [Improve performance of split\_scalar\_matrix by blegat · Pull Request #136 · jump-dev/MosekTools.jl · GitHub](https://github.com/jump-dev/MosekTools.jl/pull/136)

---

<div class="post-metadata">

**Author:** ![blegat](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/blegat/32/217090_2.png) [@blegat](https://discourse.julialang.org/u/blegat)\
**Post date:** [December 17, 2024, 12:52pm UTC](https://discourse.julialang.org/t/updating-to-mosektoolsv0-15-3-breaks-many-jump-sdp-codes/123893/5 "2024-12-17T12:52:46Z")

</div>

@Shuvomoy_Das_Gupta Thanks for reporting it ! It should be fixed in v0.15.4 of MosekTools

---

<div class="post-metadata">

**Author:** ![Shuvomoy\_Das\_Gupta](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/shuvomoy_das_gupta/32/10069_2.png) [@Shuvomoy\_Das\_Gupta](https://discourse.julialang.org/u/Shuvomoy_Das_Gupta)\
**Post date:** [December 17, 2024, 1:09pm UTC](https://discourse.julialang.org/t/updating-to-mosektoolsv0-15-3-breaks-many-jump-sdp-codes/123893/6 "2024-12-17T13:09:37Z")

</div>

Great, thanks @blegat and @odow !