flopscope.accounting.svdvals_cost
flopscope.accounting.svdvals_cost(m, n, k=None)[flopscope source]
Weighted FLOP cost of computing singular values (values-only SVD).
Delegates to flopscope._flops.svd_cost with with_vectors=False:
2*a*b^2 + 2*b^3, a = max(m, n), b = min(m, n). Top-k (1 <= k < min(m, n))
bills min(4*m*n*k, that cost).
Parameters
- m:int
Number of rows in the input matrix.
- n:int
Number of columns in the input matrix.
- k:int | None, optional
Target rank or number of singular components to estimate. Defaults to
None.
Returns
- :int
Weighted public cost estimate, floored to match runtime accounting.
Notes
This helper multiplies the analytical FLOP count by the active weight from flopscope._weights and then applies int(...) so public estimates match budget deductions.