def test_rfftn_numpy(self):
"""Test that rfftn_numpy works as expected"""
axes = [(0, 1, 2), (0, 2, 1), (1, 0, 2), (1, 2, 0), (2, 0, 1), (2, 1, 0)]
for x in [self.ad, self.af]:
for a in axes:
r_tol, a_tol = _get_rtol_atol(x)
rfft_tr = mkl_fft.rfftn_numpy(np.transpose(x, a))
tr_rfft = np.transpose(mkl_fft.rfftn_numpy(x, axes=a), a)
assert_allclose(rfft_tr, tr_rfft, rtol=r_tol, atol=a_tol)
def _rfft2(a, overwrite_x=False, extra={}):
libname = CDDMConfig["rfft2lib"]
cutoff = a.shape[-1] // 2 + 1
if libname == "mkl_fft":
return mkl_fft.rfftn_numpy(a.real, axes=(-2, -1), **extra)
#return mkl_fft.fft2(a, overwrite_x = overwrite_x)[...,0:cutoff]
elif libname == "scipy":
return spfft.fft2(a, overwrite_x=overwrite_x, **extra)[..., 0:cutoff]
elif libname == "numpy":
return np.fft.rfft2(a.real,
**extra) #force real in case input is complex
elif libname == "pyfftw":
return fftw.numpy_fft.rfft2(
a.real, **extra) #force real in case input is complex
def _rfft2(a, overwrite_x=False, extra={}):
libname = CDDMConfig["rfft2lib"]
cutoff = a.shape[-1] // 2 + 1
if libname == "mkl_fft":
out = mkl_fft.rfftn_numpy(a.real, axes=(-2, -1), **extra)
#return mkl_fft.fft2(a, overwrite_x = overwrite_x)[...,0:cutoff]
elif libname == "scipy":
out = spfft.fft2(a, overwrite_x=overwrite_x, **extra)[..., 0:cutoff]
elif libname == "numpy":
out = np.fft.rfft2(a.real,
**extra) #force real in case input is complex
elif libname == "pyfftw":
out = fftw.numpy_fft.rfft2(
a.real, **extra) #force real in case input is complex
# depending on how the libraries are compiled, the output may not be of same dtype as requested
# float32 may be converted to complex128... so we make sure it is of specified type.
return np.asarray(out, CDTYPE)