FFT vs Huygens notebook uses old API

[dkirkby]

This notebook uses old APIs for both PSF methods. I think I fixed the first off-axis example with the following changes, but please double check @jmeyers314:

#theta_y = 0.75
theta_y = np.deg2rad(0.75)
...
fftpsf = batoid.fftPSF(telescope, theta_x, theta_y, wavelength)
fftpsf, coords = fftpsf.array, fftpsf.coords
extent = np.array([coords[0,0,0], coords[-1,-1,0], coords[0,0,1], coords[-1,-1,1]])
#scale, fftpsf = batoid.fftPSF(telescope, wavelength, theta_x, theta_y)
fftpsf = fftpsf[16:-16, 16:-16]
#extent = scale*fftpsf.shape[0]/2*np.r_[-1., 1., -1., 1.] # radians
#extent -= scale/2
...
#goodRays = batoid.trimVignetted(traced_rays)
goodRays = traced_rays.trimVignetted()
...
#huygensPSF = batoid.huygensPSF(telescope, xs=xs, ys=ys, zs=None, rays=orig_rays, saveRays=False)
huygensPSF = batoid.huygensPSF(telescope, theta_x, theta_y, wavelength, nx=16)
huygensPSF = huygensPSF.array

The resulting images are: