The formulas given in the MLS paper are presumably written-out expressions for the direction cosines indicated, but result in different values than simply computing dot products between the vectors ei and the input diffraction vector.
I'm unclear on what they are doing in eqns 3-7, since a direct transcription of of the equations seems to be producing erroneous results.
The NANs were arising from sqrt(1 - tau_2**2 - tau_3**2), where the later terms summed to values larger (in magnitude) than 1.
The formulas given in the MLS paper are presumably written-out expressions for the direction cosines indicated, but result in different values than simply computing dot products between the vectors ei and the input diffraction vector.
I'm unclear on what they are doing in eqns 3-7, since a direct transcription of of the equations seems to be producing erroneous results.
The NANs were arising from
sqrt(1 - tau_2**2 - tau_3**2)
, where the later terms summed to values larger (in magnitude) than 1.