clamp in C51

[XinJingHao]

Hi! The repo C51 really helps a lot!

However, I have a question. In line 226 and 227:
l = b.floor().clamp(0, args.n_atoms - 1)
u = b.ceil().clamp(0, args.n_atoms - 1)

It seems that the clamp() function is redundant? Because t_z is already in v_min and v_max.
Why do we still use it here?

Looking forward to your reply~
Thanks