Open giadarol opened 4 years ago
@giadarol : would it be acceptable to use
(x*x + y*y) < ( min_sigma_diff * min_sigma_diff)
instead of
(fabs(x) + fabs(y)) < min_sigma_diff
This would be easier to vectorize and we can reuse the x*x + y*y
term multiple times in the same function. Plus, if I understand correctly, this is condition gauges the "real" (euclidian) distance from the reference particle rather than the maximum norm
That’s perfectly fine
Found in pysixtrack tests that beam beam has a division by zero if:
Fixed in pysixtrack, implement also in Sixtracklib