User Guide

Core workflow

1. Build a SparseGrid(dim, level).
2. Generate points with generate_points().
3. Set nodal values in each point's fv.
4. Convert nodal values using nodal_2_hier().
5. Evaluate with eval_funct(x).

Main objects

• SparseGrid
• Manages sparse-grid structure and transforms.
• GridPoint
• Stores point coordinates (pos), nodal value (fv), and hierarchical value (hv).

Data layout

• indices: list of interleaved indices [l_1, p_1, ..., l_d, p_d].
• g_p: dictionary keyed by index tuple, value is GridPoint.
• domain: tuple of intervals per dimension, default [0, 1]^d.

Typical usage notes

• Always call generate_points() before filling values.
• Always call nodal_2_hier() before eval_funct().
• Use consistent dimension between x and the grid dim.