refpts offset for acceptance

[swhite2401]

This PR answers #637. Individual offset can now be input for each refpts in get_acceptance in the form of an ndarray.

[swhite2401]

@oscarxblanco, everything seems to work fine on my side could you please let me know if this is what you were looking for?

[oscarxblanco]

@swhite2401 I have made several tests to obtain the momentum aperture :

1. using get_momentum_acceptance and passing the closed orbit and their reference points.
2. using get_acceptance and passing the closed orbit and their reference points
3. using get_acceptance and passing only the closed orbit.

All three finish without issues. The lifetime calculated from these three methods differs by less than 1%. The reason of this difference is not completely clear, but in any case, the modifications you made are OK.

I only have a doubt about a line that is removed where the variable 'istracked' is not longer assigned to True. It seems to have no impact but its name seems relevant.

[swhite2401]

Yes I saw this small difference as well, my interpretation is that in one case you provide the full closed orbit looking with the default ring starting point and in the other case you rotate the ring and recompute the closed orbit for each repfts... it could be that the 6d orbit converges to slightly different values.

Concerning istracked pycharm complained about this line and in fact it is ok because this varaible is overwritten in the while loop and not used anywhere else so no need to initialize outside the loop.

[oscarxblanco]

It's OK on my side. Thank you for this modifications !