Available plugins¶
gemseo-benchmark¶
benchmark
This package provides functionalities to benchmark algorithms, that is, to measure and compare their performances from a catalog of problems. The result is a benchmarking report in HTML or PDF format illustrated with data profiles.
gemseo-bilevel-outer-approximation¶
optimization
This package provides the bi-level outer approximation (shortened as bi-level OA) algorithm to solve mixed integer optimization problems. This algorith has proven itself for structural optimization problems in particular.
gemseo-calibration¶
calibration
This package provides functionalities for calibrating inputs of disciplines from a set of observations of other inputs and outputs, using error metrics and optimizers. These approaches can be applied to a single discipline, or to a set of loosely or tightly coupled disciplines.
gemseo-fmu¶
FMU co-simulation
This package provides functionalities for loading, interacting, and co-simulating Functional Mockup Unit models (FMUs). It enables the interoperability of FMUs with the different features of GEMSEO, e.g. multidisciplinary design optimization and uncertainty quantification.
gemseo-hexaly¶
optimization
This package provides an interface to the Hexaly hybrid optimization solver.
gemseo-http¶
remote HTTP
This package exposes GEMSEO disciplines as RESTful web services. It bridges the gap between local MDO processes and remote computing resources by providing a seamless client-server interface.
gemseo-jax¶
discipline JAX
This package provides functionalities to accelerating MDO by leveraging the power of JAX (a Python library for high-performance array computing). Among them, there are a JAX discipline wrapping a JAX function and a chain of JAX disciplines to avoid JAX-to/from-NumPy conversions.
gemseo-matlab¶
discipline Matlab
This package provide a discipline able to wrap any Matlab code defined as a .m file and use it as any other discipline. It can be used on encrypted, MATLAB built-in and user functions.
gemseo-mlearning¶
machine learning
This package provides an active learning algorithm to leverage surrogate models, notably Gaussian process regressors, in order to sequentially estimate various quantities of interest (optimas, quantiles, contours, expected values, etc.). It is compatible with any surrogate provider.
gemseo-mma¶
optimization MMA
This package provides an implementation of the MMA (Method of Moving Asymptotes) optimization algorithm. This algorithm can be used for single objective continuous optimization problems with non-linear inequality constraints.
gemseo-pdfo¶
optimization PDFO
This package provides an interface to the PDFO library, including derivative-free optimization algorithms. plugin for the PDFO library. The available optimizers are BOBYQA, COBYLA and NEWUOA.
gemseo-petsc¶
linear solver PETSc
This package provides an interface to the PETSc linear solvers and ordinary differential equations (ODE) solvers. Linear solvers can be used for direct and adjoint linear system resolution in GEMSEO. The ODE solver provides the computation of the adjoints with respect to the initial conditions of the ODE and with respect to the design variables.
gemseo-pseven¶
optimization pseven
This package provides an interface to the pSeven library for optimization.
gemseo-pymoo¶
optimization pymoo
This package provides an interface to the pymoo library for multi-objective optimization. The available algorithms are NSGA-II, NSGA-III, R-NSGA-III and U-NSGA-III. It also interfaces visualization techniques proposed by pymoo.
gemseo-pyoptsparse¶
optimization pyoptsparse
This package provides an interface to the pyOptSparse library for nonlinear constrained optimization. The available algorithms are SLSQP and SNOPT.
gemseo-scilab¶
discipline Scilab
This package provide a discipline able to wrap any Scilab function code defined in a .sci file and use it as any other discipline in a multidisciplinary study.
gemseo-ssh¶
remote SSH
This package allows to delegate the execution of a discipline or any sub-process to a (such as an MDA or MDOScenario, or MDOChain) to a remote machine via SSH. It allows you to distribute MDO workflows across multiple machines and multiple systems (Linux, Windows, MacOS). It can be combined with GEMSEO's job scheduler interface to send disciplines to a remote HPC and add them to the job scheduler queue.
gemseo-template-editor-gui¶
discipline GUI
This package provides a graphical user interface to create input and output file templates for the DiscFromExe discipline.
gemseo-umdo¶
uncertainty quantification MDO
This package provides functionalities to solve multidisciplinary optimization problems under uncertainty, by combining MDO formulations and statistics estimation techniques. It also proposes advanced techniques for uncertainty quantification and management, including variance reduction techniques and visualization tools.