NumSharp.Interop.pythonnet 0.81.0

PyCmaSampler for OptunaSharp — a native OptunaSharp BaseSampler that drives the real pycma (Nikolaus Hansen's `cma` package: cma.CMAEvolutionStrategy + cma.BoundTransform) library in-process via Python.NET (pythonnet). A faithful C# port of optuna_integration's PyCmaSampler glue (the sampler protocol + the parameter/cma-space transforms + the ask/tell generation replay) whose CMA-ES numerics are direct pycma calls, so it is byte-exact with pycma. Distinct from OptunaSharp's native CmaEsSampler, which ports the separate `cmaes` package — this mirrors Optuna's own optuna.integration.PyCmaSampler vs optuna.samplers.CmaEsSampler split. Does NOT require the Python optuna_integration package — install only `cma`. The .NET analog of using optuna_integration.PyCmaSampler, built on OptunaSharp.Interop.pythonnet. Part of the SciSharp STACK.

PymooSampler for OptunaSharp — a native OptunaSharp BaseSampler that drives the real pymoo (the multi-objective evolutionary-optimization library) in-process via Python.NET (pythonnet). One sampler exposes pymoo's whole multi-objective algorithm zoo — NSGA-II, NSGA-III, R-NSGA-II, R-NSGA-III, U-NSGA-III, AGE-MOEA, AGE-MOEA-2, RVEA, SMS-EMOA, SPEA2, C-TAEA — through a pythonnet-free PymooAlgorithm enum. The Optuna side (study, storage, ask/tell, search space) is native OptunaSharp; the operators, reference-direction machinery and HV/decomposition survival are genuine pymoo, reconstructed from the study's completed trials and driven through pymoo's native ask()/tell() — the same reconstruct-from-trials bridge as OptunaSharp.Interop.PyCma / .BoTorch. Configurable through PymooAlgorithmOptions: SBX crossover + PM mutation, duplicate elimination, RVEA alpha/adapt_freq, R-NSGA weights/mu/epsilon, and Das-Dennis / Riesz-energy reference directions. Supports constrained optimization (pymoo's G ≤ 0 feasibility, via Trial.SetConstraint or a constraintsFunc) and single-, two- and many-objective studies over Optuna's full parameter space (float/int with log/step, categoricals, dynamic spaces). Beyond Optuna core (Optuna ships only NSGA-II/III), a live-runtime complement to the native AgeMoea2Sampler. Does NOT require any Python glue package — install only `pymoo`. Built on OptunaSharp.Interop.pythonnet. Part of the SciSharp STACK.

BoTorchSampler for OptunaSharp — a native OptunaSharp BaseSampler that drives the real BoTorch (PyTorch + GPyTorch) library in-process via Python.NET (pythonnet). A faithful C# port of optuna_integration's BoTorchSampler glue (sampler protocol + candidates_func selection) whose numerics are direct botorch calls, so it is byte-exact with BoTorch. You pick the acquisition function with the pythonnet-free BoTorchCandidatesFunc enum: LogEI, qEI, qLogEI and qNEI for one objective; qEHVI, EHVI, qNEHVI, qParEGO and qLogEI-ParEGO for several; the knowledge-gradient qKG and qHVKG; constrained optimization through Trial.SetConstraint. Does NOT require the Python optuna_integration package — install only `botorch`. The .NET analog of using optuna_integration.BoTorchSampler, built on OptunaSharp.Interop.pythonnet. Part of the SciSharp STACK.