Unfolding with non-square A

 Claim: Using non-square A doesn't utilise embedding statistics efficiently.

Compare unfolding with 8 particle and 8 detector bins to unfolding with 8 particle and 11 detector bins.

Particle bins (in both cases): [ 10. 12.3 15. 18.3 22.4 27.4 33.5 40.9 50. ]
8 detector bins [ 11.5 13.25 15.5 18.5 22.2 27.0 33.0 40.9 50.8 ]
11 detector bins [ 11.5 12.7 14.3 16. 18.2 20.8 23.8 27.5 31.9 37.1 43.4 50.8 ]

The unfolding was performed without regularization.

The numerical errors relative to embedding reference are
in 8x8 case: 0.058, 0.038, 0.03 , 0.026, 0.026, 0.04 , 0.048, 0.086
in 11x8 case: 0.065, 0.045, 0.043, 0.053, 0.075, 0.139, 0.332, 0.746