Optimization Solver:Download historical versions

Download V1.1.1

Release date: 2024.03.06

Changelog:

V1.1.1 Release date: March 6, 2024

• Fixed one bug.

V1.1.0 Released: January 30, 2024

• Improved MILP and QP algorithm performance.

• Added five tunable MILP parameters.

• Added the SolutionNumber parameter to retrieve multiple suboptimal solutions during MIP solving.

• Added the C# API with corresponding documentation and examples.

• Added support for calling MindOpt from Julia using JuMP and AmplNLWriter.jl.

• Improved the license validation scheme.

• Fixed bugs:

  • Fixed a precision bug that occurred when writing LPF files.

  • Fixed incorrect solve time reporting during parallel solving.

  • Fixed an issue where all solve tasks terminated due to iteration limits during parallel solving.

  • Fixed bugs related to simplex method warm starts.

  • Fixed a bug in dense column count statistics during QP solving.

  • Fixed a preprocessing method bug during QP solving.

Download V1.0.1

Release date: 2023.11.14

Changes:

• Improved algorithm performance and fixed bugs.

• Added a new license type to support custom-duration solver authorizations purchased on Alibaba Cloud.

Download V1.0.0

Release date: 2023.10.12

Changes:

• Introduced a brand-new MindOpt SDK API and corresponding code examples. This version is incompatible with V0.x interfaces.

• Added MILP warm start, SOS constraints, and indicator constraints.

• Added MILP IIS solving capability and upgraded the LP IIS solving module.

• Added callback function support.

• Added support for the .qps format in input files.

• Improved the algorithm workflow of concurrent optimization methods.

• Provided primal-dual feasible solutions for linear programming problems.

• Improved algorithm performance and fixed bugs.

Download V0.25.1

Release date: 2023.08.16

V0.25.1 Release date: August 16, 2023

Fixed one bug.

V0.25.0 Publish date: August 8, 2023

Changes:

• Improved the performance and stability of the interior point method.

• Improved Python API performance.

• Improved IIS algorithm stability.

• Added the MILP IIS algorithm.

• API changes:

  • New error codes:

    • MDO_INVALID_ARGUMENT(-3), MDO_IIS_NUMERIC(-30000), MDO_IIS_NO_SOLN(-3004), and MDO_IIS_FEASIBLE(-3005).

  • New attributes:

    • C API: MDO_INT_ATTR_ROW_IIS and MDO_INT_ATTR_COL_IIS.

    • C++ API: MDO_INT_ATTR::ROW_IIS and MDO_INT_ATTR::COL_IIS.

    • Python API: MDO_INT_PARAM.ROW_IIS and MDO_INT_ATTR.COL_IIS.

    • JAVA API: INT_ATTR::ROW_IIS and INT_ATTR::COL_IIS.

  • Updated APIs:

    • C API: Mdo_computeIIS.

    • C++ API: MdoModel::computeIIS.

    • Python API: MdoModel.compute_iis.

    • JAVA API: MdoModel::computeIIS.

  • New integer parameters:

    • C API: MDO_INT_PARAM_MIP_MAX_STALLING_NODES, MDO_INT_PARAM_MIP_MAX_SOLS, MDO_INT_PARAM_MIP_DETECT_DISCONNECTED_COMPONENTS, MDO_INT_PARAM_MIP_SOLUTION_POOL_SIZE, and MDO_INT_PARAM_MIP_ALLOW_DUAL_PRESOLVE.

    • C++ API: MDO_INT_PARAM::MIP_MAX_STALLING_NODES, MDO_INT_PARAM_MIP_MAX_SOLS, MDO_INT_PARAM::MIP_DETECT_DISCONNECTED_COMPONENTS, MDO_INT_PARAM::MIP_SOLUTION_POOL_SIZE, and MDO_INT_PARAM::MIP_ALLOW_DUAL_PRESOLVE.

    • Python API: MDO_INT_PARAM.MIP_MAX_STALLING_NODES, MDO_INT_PARAM.MIP_MAX_SOLS, MDO_INT_PARAM.MIP_DETECT_DISCONNECTED_COMPONENTS, MDO_INT_PARAM.MIP_SOLUTION_POOL_SIZE, and MDO_INT_PARAM.MIP_ALLOW_DUAL_PRESOLVE.

    • JAVA API: INT_PARAM::MIP_MAX_STALLING_NODES, INT_PARAM_MIP_MAX_SOLS, INT_PARAM::MIP_DETECT_DISCONNECTED_COMPONENTS, INT_PARAM::MIP_SOLUTION_POOL_SIZE, and INT_PARAM::MIP_ALLOW_DUAL_PRESOLVE.

  • New real number parameters:

    • C API: MDO_REAL_PARAM_MIP_LINEARIZATION_BIG_M and MDO_REAL_PARAM_MIP_CUTOFF.

    • C++ API: MDO_REAL_PARAM::MIP_LINEARIZATION_BIG_M and MDO_REAL_PARAM::MIP_CUTOFF.

    • Python API: MDO_REAL_PARAM.MIP_LINEARIZATION_BIG_M and MDO_REAL_PARAM.MIP_CUTOFF.

    • JAVA API: REAL_PARAM::MIP_LINEARIZATION_BIG_M and REAL_PARAM::MIP_CUTOFF.

• Fixed bugs:

  • Fixed SDK-related issues.

  • Fixed performance issues that occurred when deleting constraints.

  • Fixed issues that occurred when reading MPS/LP files.

Download V0.24.1

Release date: 2023.04.23

Changes:

• Added Linux ARM and Apple Silicon installation packages.

• Added pip installation support for the Python library: pip install mindoptpy.

• Improved MILP solver performance.

• Renamed the data masking interface names to --sanitize and --undo_sanitize. For more information, see Data masking.

Download V0.24.0

Release date: 2023.03.30

Changes:

• Improved MILP solver performance. This is an official release. All users on Alibaba Cloud with authenticated accounts now have full access and do not need to apply for special licenses.

• Improved concurrent optimization method performance.

• Redesigned the Cholesky factorization code in IPM.

• API changes:

  • Added a command-line feature for data masking. For more information, see Data masking:

    • --desensitize masks model files by removing business-sensitive data, such as variable and constraint names, leaving only numerical data that reveals no business context.

    • --sensitize restores masked files.

  • New integer parameters:

    • C API: MDO_INT_PARAM_MIP_AUTO_CONFIGURATION.

    • C++ API: MDO_INT_PARAM::MIP_AUTO_CONFIGURATION.

    • Python API: MDO_INT_PARAM.MIP_AUTO_CONFIGURATION.

    • JAVA API: MDO_INT_PARAM::MIP_AUTO_CONFIGURATION.

  • New error codes:

    • MDO_MODEL_INVALID_CHAR_ATTR(-2014) and MDO_MODEL_INVALID_ATTR_NAME(-2019).

• Fixed bugs:

  • Fixed a bug in the Mdo_loadProb API.

  • Fixed bugs in the LP writer and MPS writer, including issues that occurred when writing GZ/BZ2 compressed formats.

  • Fixed a bug related to Mdo_copyMdl.

  • Fixed a minor bug in the AMPL application (mindoptampl).

Download V0.23.1

Release date: 2022.12.21

Changes:

• This is the macOS M1 chip version compiled based on V0.23.0. It does not include MILP solving and supports only LP, QP, and SDP solving.

Download V0.23.0

Release date: 2022.11.29

Changes:

• Added semidefinite programming (SDP) related APIs. Added support for the dat-s format in input files.

• Added documentation for the .nl format. The examples folder in the installation package contains .nl file examples. You can use the command mindoptampl filename.nl to call the solver directly.

• Fixed a minor bug in the LPF reader.

• Fixed several MILP solving bugs and optimized MILP performance. (Note: The MILP feature is currently available only with limited test licenses. Please email us to apply.)

• Added new APIs for C, C++, Python, and Java, such as the Python API: MdoModel.add_sym_mat, MdoModel::add_sym_mats, MdoModel::replace_sym_mat_objs, MdoModel::replace_sym_mat_elements, MdoModel::get_real_attr_sym_mat, and MdoModel::copy_model.

• Added new error codes:

  • MDO_MODEL_INVALID_SYM_MAT_IDX (-2006), MDO_MODEL_INVALID_SYM_MAT_ROW_IDX (-2007), MDO_MODEL_INVALID_SYM_MAT_COL_IDX (-2008), and MDO_MODEL_INVALID_REAL_ATTR_SYM_MAT (-2013).

• Modified error codes:

  • MDO_LIB_FOLDER_NOT_FOUND (-12) is replaced by MDO_DLL_ERROR (-12).

• Added new attributes for C, C++, Python, and Java, such as the Python API: MDO_REAL_ATTR.SYM_MAT_PRIMAL_SOLN, MDO_REAL_ATTR.MIP_GAP_ABS, and MDO_REAL_ATTR.MIP_GAP_REL.

Download V0.20.2

Release date: 2022.10.24

Changes:

• Fixed bugs that caused file reading failures with special lengths and issues in specific hardware environments.

Download V0.20.1

Release date: 2022.09.23

Changes:

• Optimized MILP performance and adjusted the parameter range for MIP/rootParallelism.

• Optimized the Python SDK installation script to fix errors that occurred on certain Python versions in Windows.

• Optimized the runtime software output for better clarity.

Download V0.20.0

Release date: 2022.08.31

Changes:

• Added a branch-and-cut solver for mixed integer linear programming (MILP). Added AMPL interface support for MILP problems.

• Optimized license validation with the following changes:

  • Introduced new license-free logic: Optimization problems with fewer than 300 variables and constraints can be solved without a license file (mindopt.lic or fl_client.ini).

  • Problems that exceed these limits still require the original license. To apply for authorization files, see License setup.

  • Note: Licenses for solving MILP problems are currently available only to a limited number of testers. Email us to apply, and we will manually grant permissions. For contact information, see email.

• Optimized the installation package to simplify the installation steps:

  • The installation scripts for Linux and macOS now run interactively to confirm the installation path, support custom paths, and automatically configure environment variables. For Windows, a warning is now displayed if you attempt to install the package again.

• Added multiple interfaces with MILP functionality. For more information, see the user guide.

Download V0.19.0

Release date: 2022.05.11

Changes:

• Added support for the input and solving of convex quadratic programming (QP) problems within nonlinear programming (NLP). This includes the interfaces MdoModel.set_quadratic_elements, MdoModel.get_quadratic_elements, MdoModel.delete_quadratic_elements, MdoModel.delete_all_elements, and MdoModel.delete_all_quadratic_elements.

• Improved the performance of the simplex and interior point methods. Enhanced the functionality and stability of the AMPL and concurrent methods.

• Redesigned license management with new APIs, such as the MdoEnv class, to support high-concurrency, low-latency scenarios.

• Optimized the installation methods and documentation for the installation package.

Download V0.15.1

Release date: 2021.11.08

Changes:

• Added the network flow simplex method and the multi-threaded simplex method.

  • The new network flow method shows excellent performance on the Large Network-LP Benchmark. For more information, see this article. For problems with a network structure, you can use this method by setting the new EnableNetworkFlow parameter to 1.

• Improved the stability of the simplex and interior point methods.

• Updated the data reading module.

• API changes:

  • New parameters: EnableNetworkFlow, SPX/PrimalPricing, and SPX/DualPricing.

  • Updated parameters: Presolve and Method.

  • New error codes: MDO_HOME_ENV_NOT_FOUND and MDO_LIB_FOLDER_NOT_FOUND.

  • New APIs:

    • C API: Mdo_setStrAttr and Mdo_getStrAttr.

    • C++ API: MdoModel::setStrAttr and MdoModel::getStrAttr.

    • Python API: MdoModel.set_str_attr and MdoModel.get_str_attr.

    • Java API: None. This API already existed in previous versions.

  • New model attribute: ProbName.

Download V0.14.0

Release date: 2021.09.26

Changes:

• Added support for Java APIs and Floating Licenses.

• Improved the performance of the Simplex and IPM algorithms.

• Added IIS APIs for C, C++, and Python.

• Added the Python quicksum API and optimized the Python API design.

• Added initial support and examples for the AMPL and Pyomo modeling languages.