Minor IMRPhenomXAS phase difference exists between the PyCBC and the WF4Py implementations

[chendawe]

waveform parameters:

spin1z = 0.9
spin2z = 0.9

m1 = 20
m2 = 140

distance = 421.01,
# inclination = 2.41,
inclination = 2.71,
coa_phase = 3.00,
# delta_t = -0.02247,
delta_t = 1/4096,

f_lower = 20,
delta_f = 1/16,
f_final = 2048,
f_ref=20

[chendawe]

Sorry, please let the problem pend for now. The difference could come from a fault in my code, I'm afraid.

[chendawe]

The difference seems do eappear when I try the tutorial code :

zs = np.array([.2])

events = {'Mc':np.array([30])*(1.+zs), 'dL':(Planck18.luminosity_distance(zs).value/1000.),
         'iota':np.array([.0]), 'eta':np.array([0.1]), 'chi1z':np.array([0.8]), 'chi2z':np.array([-0.8]),
          'Lambda1':np.array([0.]), 'Lambda2':np.array([0.]), 'f_lower':np.array([20]), 'f_final':np.array([2048])}

The phase difference fluctuates in the order of magnitude around 0.01

and when considering a more moderate mass ration like \eta=0.24 :

And the difference especially appears in the inspiral as far as I can see