Some questions arise about using K-SpecPart

[Panpan7411]

Hello!
The following question came up when I was running and using K-SpecPart, and I hope someone can help answer it. The version I use of julia is version 1.6.6 (2022-03-28) because I didn't find version 1.6.7.
Thanks!

julia> include("K_SpecPart/specpart.jl")
WARNING: replacing module SpecPart.
Main.SpecPart

julia> include("K_SpecPart/specpart.jl")
WARNING: replacing module SpecPart.
Main.SpecPart

julia> using Main.SpecPart

julia> SpecPart.specpart_run("benchmark/ISPD_benchmark/ibm02.hgr",hypergraph_fixed_file="",hint_file="K_specpart_solutions/ISPD_weight_benchmarks/ub_factor_2/3_way/ibm02.weight.hgr.specpart.ubfactor.2.part.3",imb=2,num_parts=3)
[ Info: ============================================================
[ Info: SpecPart K-way hypergraph partitioner**
[ Info: ============================================================
[ Info: Hypergraph info**
[ Info: ============================================================
[ Info: Num vertices 19601
[ Info: Num hyperedges 19584
[ Info: Fixed vertices 0
[ Info: ============================================================
[ Info: Post processing**
[ Info: ============================================================
[ Info: Num vertices 19601
[ Info: Num hyperedges 19584
[ Info: Hint partition cutsize 334
[ Info: ============================================================
[ Info: Solver parameters
[ Info: ============================================================
[ Info: Solver iterations 40
[ Info: Num vecs 2
[ Info: Tol 1e-40
[ Info: [specpart] iteration 1
[ Info: ============================================================
┌ Info: Results of LOBPCG Algorithm
│ * Algorithm: LOBPCG - CholQR
│ * λ: [2.0613316601657896e-9,2.1282464603512593e-9]
│ * Residual norm(s): [3.425383452041066e-8,4.4027298331682545e-7]
│ * Convergence
│ * Iterations: 41
│ * Converged: false
└ * Iterations limit: 40
[ Info: ============================================================
[ Info: ============================================================
┌ Info: Results of LOBPCG Algorithm
│ * Algorithm: LOBPCG - CholQR
│ * λ: [4.895575495828814e-10,1.4156011362928634e-9]
│ * Residual norm(s): [3.1412502093481887e-7,3.867865718074016e-7]
│ * Convergence
│ * Iterations: 41
│ * Converged: false
└ * Iterations limit: 40
[ Info: ============================================================
[ Info: ============================================================
┌ Info: Results of LOBPCG Algorithm
│ * Algorithm: LOBPCG - CholQR
│ * λ: [4.88796704529273e-10,1.4333586489649734e-9]
│ * Residual norm(s): [3.4242571434758896e-7,3.606249727286522e-7]
│ * Convergence
│ * Iterations: 41
│ * Converged: false
└ * Iterations limit: 40
[ Info: ============================================================
[ Info: Using eigenvectors [1]
[ Info: Cutsize from tree sweep 217.0
[ Info: Cutsize from tree sweep 231.0
[ Info: Cutsize from tree sweep 426 with balance [6148, 6916, 6537]
ERROR: UndefVarError: neighbors not defined
Stacktrace:
[1] build_metis_graph(tree::SimpleWeightedGraphs.SimpleWeightedGraph{Int64, Int64}, metis_opts::Int64)
@ Main.SpecPart ~/project/SpecPart/HypergraphPartitioning-main/K_SpecPart/metis.jl:90
[2] METIS_tree_partition(T::Graphs.SimpleGraphs.SimpleGraph{Int64}, distilled_cuts::Main.SpecPart.cut_profile, hgraph::Main.SpecPart.hypergraph, seed::Int64, metis_path::String, metis_opts::Int64, num_parts::Int64, ub_factor::Int64)
@ Main.SpecPart ~/project/SpecPart/HypergraphPartitioning-main/K_SpecPart/tree_partition.jl:323
[3] tree_partition(adj::SparseArrays.SparseMatrixCSC{Float64, Int64}, X::Matrix{Float64}, hgraph::Main.SpecPart.hypergraph, fixed_vertices::Main.SpecPart.pindex, ub_factor::Int64, capacities::Vector{Int64}, metis_path::String, num_parts::Int64, seed::Int64, kway::Bool)
@ Main.SpecPart ~/project/SpecPart/HypergraphPartitioning-main/K_SpecPart/tree_partition.jl:506
[4] k_way_spectral_refine(hypergraph_file::String, partition::Vector{Int64}, hgraph::Main.SpecPart.hypergraph, metis_path::String, ub_factor::Int64; num_parts::Int64, eigen_vecs::Int64, cycles::Int64, refine_iters::Int64, solver_iters::Int64, best_solns::Int64, seed::Int64)
@ Main.SpecPart ~/project/SpecPart/HypergraphPartitioning-main/K_SpecPart/specpart.jl:283
[5] specpart_run(hypergraph_file::String; hypergraph_fixed_file::String, hint_file::String, imb::Int64, num_parts::Int64, eigvecs::Int64, refine_iters::Int64, solver_iters::Int64, best_solns::Int64, ncycles::Int64, seed::Int64)
@ Main.SpecPart ~/project/SpecPart/HypergraphPartitioning-main/K_SpecPart/specpart.jl:478
[6] top-level scope
@ REPL[3]:1

[bodhi91]

Hi Panpan

The error here is that SimpleWeightedGraphs.neighbors cannot be identified by Julia v1.6.6.

Version 1.6.7 is available here.

Please try this and let me know if the error still persists.

Thanks
-- Bodhi

[bodhi91]

Closing this issue.