How-To Guides

Use these pages when you already know the task and need the concrete API pattern.

Build And Update Interpolants

• Pre-computed Values builds an approximation from an existing Chebyshev grid.
• Error-Driven Construction increases node counts until a target error is reached.
• Adaptive Refinement explains coefficient-tail error checks, Sobol indices, and refinement helpers.
• Serialization & Construction saves, loads, and rebuilds approximation objects.
• Parallel Build & Progress controls multi-threaded build execution and progress reporting.

Model Difficult Functions

• Piecewise Chebyshev (Splines) places knots at known kinks or singularities.
• Special Points routes known trouble points into piecewise construction.
• Sliding Technique uses grouped low-dimensional slices for high-dimensional models with limited cross-group interaction.
• Tensor Train uses TT-Cross for high-dimensional models with broader variable coupling.
• Continuous-State Dynamic Programming shows the finite-horizon Bellman-collocation pattern.

Compute Derived Quantities

• Computing Greeks evaluates derivatives for sensitivity analysis.
• Calculus covers differentiation and integration operations.
• Chebyshev Algebra composes, adds, and multiplies approximations.
• Extrusion & Slicing fixes or adds coordinates to reshape a model.
• Advanced Usage collects lower-level usage patterns and caveats.