approximate_double

Single-threaded Dirac-to-Dirac reduction.

This implementation performs the complete optimization in a single thread.

Recommended for:

• Small to medium problem sizes
• Debugging
• Deterministic performance benchmarking

Parameters

• y: Input Dirac locations (M x N)
• M: Number of input components
• L: Number of reduced components
• N: Dimension
• x: Output buffer (L x N)
• wX: Optional reduced weights
• wY: Optional input weights

Example

d2d = dgs.DiracToDiracApproximation()
result = d2d.approximate_double(y, M, L, N, x)
 
if result.result:
    print("Optimization successful")
    print(result.x)