seemps.analysis.comptree.BranchNode#
- class seemps.analysis.comptree.BranchNode#
Bases:
object- compute_image(values: Vector, default_tol: float = 1e-4, tol_multiplier: float = 1.25) Vector[source]#
Compute the image of the node function over a set of input values.
The node function x_out = func(x_in, x_s) is evaluated for all combinations of input values and grid indices, and the resulting outputs are collected into a set of sorted unique values. The image can be optionally compressed by binning nearby values and by enforcing max_rank, which prevents a combinatorial growth of image sizes across the tree and bounds the bond dimension of the corresponding MPS core.