00:37:04	Gagan Mohanty:	Nothing happended — I guess, you mean to say “entirely nonresonant production” of three daughter particle?
00:42:25	Christoph Hanhart:	What was meant was the absence of all final state interactions - those are not necessarily resonant. But indeed: To get pronounced structures into the Dalitz plot resonances are needed.
01:02:18	Agnes Roberts:	pretty cool
01:41:23	Kirill Chilikin:	Widths of the resonances with 3-body decays depend on their decay amplitudes. One has to do numeric integration at every function calculation...
01:42:15	Jonas Rademacker:	@Kirill: yes, but this can be done in advance of your fit.
01:42:52	Jonas Rademacker:	So you do it once (unless you vary the resonance parameters of the 3 body decays, then it will be computationally very expensive.
01:43:24	Jonas Rademacker:	… in this context, there is a move towards doing amplitude fits on GPUs using CUDA or Tensorflow.
01:45:19	Gagan Mohanty:	Yes, I agree with Christoph. It is the pole of Leagender’s polynomial
01:49:15	Kirill Chilikin:	Indeed, I have to use explicit CPU vectorization (!$OMP SIMD), specialized GPUs are even better.
02:01:38	Jonas Rademacker:	Thanks for the invite!
02:02:09	Jonas Rademacker:	Bye bye!