Fix typo in QP problem instantiation #758

mtanneau · 2026-01-10T04:13:49Z

Description

Fixes #757 by removing extra argument when creating QP problem instance.

Note that since the QP problem constructor does not accept variable types, one can remove the definition of variable types which are unused. Happy to do that if preferable :)

Issue

closes #757

Checklist

I am familiar with the Contributing Guidelines.
Testing
- New or existing tests cover these changes
- Added tests
- Created an issue to follow-up
- NA
Documentation
- The documentation is up to date with these changes
- Added new documentation
- NA

Summary by CodeRabbit

Release Notes

New Features
- Added Quadratic Programming (QP) solver support alongside existing LP and MILP capabilities.
- Expanded branch-and-bound solver with multiple diving heuristic strategies for improved optimization.
Improvements
- Enhanced solver performance and stability through GPU memory optimizations and improved stream handling.
- Simplified solver configuration by removing deprecated tolerance and mode parameters.
Documentation
- Updated documentation to reflect QP solver availability across C, Python, and server APIs.

_{✏️ Tip: You can customize this high-level summary in your review settings.}

copy-pr-bot · 2026-01-10T04:13:52Z

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

coderabbitai · 2026-01-10T04:14:32Z

Caution

Review failed

Failed to post review comments

📝 Walkthrough

Walkthrough

This PR introduces new diving heuristics for branch-and-bound solving, refactors the dual simplex solver to use device-resident data, adds Quadratic Programming support to documentation, upgrades GitHub Actions workflows, updates Python APIs by removing deprecated parameters, and adds S3-based dataset downloads.

Changes

Cohort / File(s)	Summary
GitHub Actions Workflow Updates `.github/workflows/build.yaml`, `build_images.yaml`, `pr.yaml`, `test.yaml`, `trigger-breaking-change-alert.yaml`	Replaced CUDA-specific version tags with `main` branch reference; added CUOPT S3 dataset secrets to test jobs; updated copyright years and added COMMIT_TIME handling in docker image build.
Dual Simplex Core Solver Refactoring `cpp/src/dual_simplex/barrier.cu`, `barrier.hpp`, `branch_and_bound.cpp`, `branch_and_bound.hpp`, `iterative_refinement.hpp`	Migrated barrier solver to device-resident augmented matrices; replaced CPU-focused methods with GPU-centric implementations; refactored branch-and-bound with new diving strategies (pseudocost, line-search, guided, coefficient); replaced `mip_exploration_status_t` with `bnb_worker_type_t` enum; introduced `bnb_stats_t` for thread-safe statistics.
Diving Heuristics Implementation `cpp/src/dual_simplex/diving_heuristics.cpp`, `diving_heuristics.hpp`	New module implementing five diving strategies via template functions; introduced `branch_variable_t` struct; includes pseudocost, line-search, guided, and coefficient-based variable selection with logging.
Node Queue and Queue Management `cpp/src/dual_simplex/node_queue.hpp`, `diving_queue.hpp`	Removed old `diving_queue.hpp`; introduced new `node_queue_t<i_t, f_t>` with dual heaps (best-first and diving) and mutex-protected operations; added generic `heap_t<T, Comp>` template.
Basis Repair and Updates `cpp/src/dual_simplex/basis_solves.cpp`, `basis_solves.hpp`, `basis_updates.cpp`, `basis_updates.hpp`, `crossover.cpp`, `phase2.cpp`, `primal.cpp`	Extended `basis_repair` and `refactor_basis` signatures with `lower` and `upper` bound parameters to support bound-aware nonbasic variable status selection.
Pseudo-Costs and Diving Settings `cpp/src/dual_simplex/pseudo_costs.cpp`, `pseudo_costs.hpp`, `simplex_solver_settings.hpp`	Added `obj_estimate()` method to `pseudo_costs_t`; introduced `diving_heuristics_settings_t` struct with diving configuration; renamed `num_bfs_threads` to `num_bfs_workers`.
Stream and Device Code Updates `cpp/src/linear_programming/pdhg.cu`, `pdlp.cu`, `initial_scaling.cu`, `termination_strategy/*.cu`, `cython_solve.cu`, `cusparse_view.cu`	Replaced `stream_view_` with `stream_view_.value()` for raw CUDA stream handles; added runtime cuSPARSE version checking; reset stream contexts for result buffers.
Iterative Refinement Device Migration `cpp/src/dual_simplex/iterative_refinement.hpp`	Rewrote iterative refinement functions to use `rmm::device_uvector` instead of host vectors; introduced device-side functors for vector operations; changed return types to `f_t` for residual values.
Quadratic Objective and Problem Setup `cpp/src/linear_programming/optimization_problem.cu`, `cpp/src/mip/problem/problem.cu`, `problem.cuh`, `problem_helpers.cuh`	Added `set_quadratic_objective_matrix` two-phase CSR construction with deduplication; introduced `substitute_variables`, `update_variable_bounds`, and `sort_rows_by_variables` public methods; added `substitution_t` struct to presolve data.
Presolve and Probing Cache Refactoring `cpp/src/mip/presolve/probing_cache.cu`, `probing_cache.cuh`, `trivial_presolve.cuh`, `trivial_presolve_helpers.cuh`	Changed `compute_probing_cache` return type from `void` to `bool`; introduced multi-threaded probing with per-thread work queues; added substitution propagation and graph sanitization; added `remap_cache_ids` parameter.
Diversity and Local Search `cpp/src/mip/diversity/diversity_manager.cu`, `cpp/src/mip/diversity/lns/rins.cu`, `recombiners/sub_mip.cuh`	Updated thread configuration to use `num_bfs_workers` and `diving_settings`; refactored probing cache integration with time budgeting; changed solution callback to include objective parameter.
Python API Deprecation Cleanup `python/cuopt_self_hosted/cuopt_sh_client/cuopt_self_host_client.py`, `python/cuopt_server/cuopt_server/utils/linear_programming/data_definition.py`, `solver.py`, `test_lp.py`	Removed `request_excess_timeout` parameter from client; deleted deprecated tolerance fields (absolute_primal, relative_dual, etc.); removed `solver_mode` and `heuristics_only` fields; updated tests with new tolerance naming.
Documentation—QP Support `README.md`, `docs/cuopt/source/*/.rst`, `docs/cuopt/source/cuopt-c/lp-qp-milp/`, `docs/cuopt/source/cuopt-python/lp-qp-milp/`	Expanded all references from "LP/MILP" to "LP/QP/MILP"; reorganized documentation structure from `lp-milp/` to `lp-qp-milp/` folders; updated API reference titles and examples.
Conda and Python Dependencies `conda/environments/*.yaml`, `python/cuopt/pyproject.toml`, `dependencies.yaml`	Added `pyyaml>=6.0.0` as runtime dependency across conda environments and Python package specs.
Docker and Dataset Handling `ci/docker/Dockerfile`, `ci/docker/test_image.sh`, `datasets/get_test_data.sh`, `datasets/linear_programming/download_pdlp_test_dataset.sh`, `datasets/mip/download_miplib_test_dataset.sh`	Added CUDA headers stage in Dockerfile; added S3 download functions with AWS credentials support; updated test harness with pytest/pexpect; added COMMIT_TIME tracking.
Utilities and Configuration `cpp/src/utilities/cuda_helpers.cuh`, `omp_helpers.hpp`, `cpp/src/mip/utilities/work_unit_ordered_queue.cuh`, `cpp/src/mip/diversity/diversity_config.hpp`	Added thread-safe dynamic shared memory caching; made `omp_atomic_t::load()` const; introduced `work_unit_ordered_queue_t` for timestamp-ordered work; increased probing cache time budgets.
Benchmarks and Configuration `benchmarks/linear_programming/cuopt/run_mip.cpp`, `benchmarks/linear_programming/utils/get_datasets.py`, `datasets/cuopt_service_data/lpmip_configs.json`	Added `mip_gap` and `is_optimal` metrics to benchmark output; enabled netlib dataset processing; removed deprecated tolerance configuration keys.
C API Examples `docs/cuopt/source/cuopt-c/lp-qp-milp/examples/simple_qp_example.c`, `docs/cuopt/source/cuopt-python/lp-qp-milp/examples/simple_qp_example.py`	Removed `variable_types` argument from QP example call; added linear constraint to Python QP example.
Routing Multi-GPU Setup `cpp/src/routing/utilities/cython.cu`	Added per-thread CUDA device context setup in batch routing loops.
QP Unit Tests `cpp/tests/qp/CMakeLists.txt`, `cpp/tests/qp/unit_tests/no_constraints.cu`	Added unconstrained QP test case covering diagonal quadratic objective solving.

Estimated code review effort

🎯 4 (Complex) | ⏱️ ~60 minutes

Possibly related PRs

Diving Improvements #697: Implements dual-heap node queue and diving heuristics infrastructure that this PR depends on; both modify branch-and-bound solver, diving strategy selection, and basis repair call sites in similar ways.

Suggested labels

improvement

Suggested reviewers

akifcorduk

🚥 Pre-merge checks | ✅ 3 | ❌ 2

❌ Failed checks (2 warnings)

Check name	Status	Explanation	Resolution
Out of Scope Changes check	⚠️ Warning	The PR contains extensive changes far beyond the scope of fixing issue `#757`, including diving heuristics, branch-and-bound refactoring, QP feature documentation, and numerous unrelated modifications.	Separate the typo fix for the QP example (the stated objective) from the large-scale architectural changes to branch-and-bound, diving heuristics, documentation updates, and other features.
Docstring Coverage	⚠️ Warning	Docstring coverage is 3.31% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The PR title is clear, specific, and directly describes the primary change: removing an extra argument from QP problem instantiation in a C example.
Linked Issues check	✅ Passed	The PR fixes issue `#757` by removing the extra variable_types argument from cuOptCreateQuadraticProblem call in the C example, resolving the compilation error.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing touches

📝 Generate docstrings

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

Signed-off-by: mtanneau <mathieu.tanneau@gmail.com>

rgsl888prabhu · 2026-01-12T15:48:46Z

Thank you @mtanneau for the PR. Will get back to you on your suggestion on variable types.

rgsl888prabhu · 2026-01-12T15:49:04Z

/ok to test f8d0336

rgsl888prabhu · 2026-01-12T15:54:59Z

Thank you @mtanneau for the PR, will ping engine team regarding your suggestion on removing.

rgsl888prabhu · 2026-01-12T15:55:37Z

/ok to test f8d0336

rgsl888prabhu · 2026-01-12T16:11:59Z

/ok to test 740b221

coderabbitai

Actionable comments posted: 0

🧹 Nitpick comments (1)

docs/cuopt/source/cuopt-c/lp-milp/examples/simple_qp_example.c (1)
105-108: Remove unused variable_types definition.

The variable_types array is now dead code since it's no longer passed to cuOptCreateQuadraticProblem. As the PR author noted, this should be removed to avoid confusing users who might wonder why the variable exists but isn't used.
♻️ Suggested fix
-  // Variable types (continuous)
-  // From the constraints:
-  // x1, x2 >= 0
-  char variable_types[] = {CUOPT_CONTINUOUS, CUOPT_CONTINUOUS};
-
   cuopt_int_t status;

📜 Review details

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 923bdf1 and 740b221.

📒 Files selected for processing (1)

docs/cuopt/source/cuopt-c/lp-milp/examples/simple_qp_example.c

🧰 Additional context used

📓 Path-based instructions (1)

docs/**/*

⚙️ CodeRabbit configuration file

docs/**/*: For documentation changes, focus on:

Accuracy: Verify code examples compile and run correctly

Completeness: Check if API changes (parameters, return values, errors) are documented

Clarity: Flag confusing explanations, missing prerequisites, or unclear examples

Consistency: Version numbers, parameter types, and terminology match code

Examples: Suggest adding examples for complex features or new APIs

Missing docs: If PR changes public APIs without updating docs, flag as HIGH priority

When code changes affect docs:

Suggest specific doc files that need updates (e.g., docs/cuopt/api.rst)

Identify outdated information contradicting the code changes

Recommend documenting performance characteristics, GPU requirements, or numerical tolerances

Files:

docs/cuopt/source/cuopt-c/lp-milp/examples/simple_qp_example.c

🧠 Learnings (13)

📓 Common learnings

Learnt from: chris-maes
Repo: NVIDIA/cuopt PR: 500
File: cpp/tests/linear_programming/c_api_tests/c_api_test.c:1033-1048
Timestamp: 2025-12-06T00:22:48.638Z
Learning: In cuOPT's quadratic programming API, when a user provides a quadratic objective matrix Q via set_quadratic_objective_matrix or the C API functions cuOptCreateQuadraticProblem/cuOptCreateQuadraticRangedProblem, the API internally computes Q_symmetric = Q + Q^T and the barrier solver uses 0.5 * x^T * Q_symmetric * x. From the user's perspective, the convention is x^T Q x. For a diagonal Q with values [q1, q2, ...], the resulting quadratic terms are q1*x1^2 + q2*x2^2 + ...

Learnt from: CR
Repo: NVIDIA/cuopt PR: 0
File: .github/.coderabbit_review_guide.md:0-0
Timestamp: 2025-11-25T10:20:49.822Z
Learning: Applies to **/*.{cu,cuh,cpp,hpp,h} : Ensure variables and constraints are accessed from the correct problem context (original vs presolve vs folded vs postsolve); verify index mapping consistency across problem transformations

Learnt from: aliceb-nv
Repo: NVIDIA/cuopt PR: 527
File: cpp/src/mip/diversity/lns/rins.cu:167-175
Timestamp: 2025-10-22T14:25:22.899Z
Learning: In MIP (Mixed Integer Programming) problems in the cuOPT codebase, `n_integer_vars == 0` is impossible by definition—MIP problems must have at least one integer variable. If there are no integer variables, it would be a pure Linear Programming (LP) problem, not a MIP problem.

📚 Learning: 2025-12-06T00:22:48.638Z

Learnt from: chris-maes
Repo: NVIDIA/cuopt PR: 500
File: cpp/tests/linear_programming/c_api_tests/c_api_test.c:1033-1048
Timestamp: 2025-12-06T00:22:48.638Z
Learning: In cuOPT's quadratic programming API, when a user provides a quadratic objective matrix Q via set_quadratic_objective_matrix or the C API functions cuOptCreateQuadraticProblem/cuOptCreateQuadraticRangedProblem, the API internally computes Q_symmetric = Q + Q^T and the barrier solver uses 0.5 * x^T * Q_symmetric * x. From the user's perspective, the convention is x^T Q x. For a diagonal Q with values [q1, q2, ...], the resulting quadratic terms are q1*x1^2 + q2*x2^2 + ...

Applied to files:

docs/cuopt/source/cuopt-c/lp-milp/examples/simple_qp_example.c