pynufft 2020.2.4

Python non-uniform fast Fourier transform (PyNUFFT)

PyNUFFT: Python non-uniform fast Fourier transform

A minimal "getting start" tutorial is available at http://jyhmiinlin.github.io/pynufft/ .

Installation

$ pip install pynufft --user

Using Numpy/Scipy

$ python
Python 3.6.11 (default, Aug 23 2020, 18:05:39) 
[GCC 7.5.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> from pynufft import NUFFT
>>> import numpy
>>> A = NUFFT()
>>> om = numpy.random.randn(10,2)
>>> Nd = (64,64)
>>> Kd = (128,128)
>>> Jd = (6,6)
>>> A.plan(om, Nd, Kd, Jd)
0
>>> x=numpy.random.randn(*Nd)
>>> y = A.forward(x)

Using PyCUDA

>>> from pynufft import NUFFT, helper
>>> import numpy
>>> A2= NUFFT(helper.device_list()[0])
>>> A2.device
<reikna.cluda.cuda.Device object at 0x7f9ad99923b0>
>>> om = numpy.random.randn(10,2)
>>> Nd = (64,64)
>>> Kd = (128,128)
>>> Jd = (6,6)
>>> A2.plan(om, Nd, Kd, Jd)
0
>>> x=numpy.random.randn(*Nd)
>>> y = A2.forward(x)

Using NUDFT_cupy and NUDFT (double precision)

Some users ask for double precision. So NUDFT and NUDFT_cupy are offered. Speed is not great though.

>>> from pynufft import NUDFT_cupy, NUDFT
>>> import numpy
>>> A2= NUDFT_cupy()
>>> om = numpy.random.randn(10,2)
>>> Nd = (64,64)
>>> A2.plan(om, Nd)
>>> x=numpy.random.randn(*Nd)
>>> y = A2.forward(x)
>>> A = NUDFT()
>>> A.plan(om, Nd)
>>> y_cpu = A.forward(x)
>>> print(numpy.linalg.norm(y.get() - y_cpu))
6.752054788357788e-14

Testing GPU acceleration

Python 3.6.11 (default, Aug 23 2020, 18:05:39) 
[GCC 7.5.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> from pynufft import tests
>>> tests.test_init(0)
device name =  <reikna.cluda.cuda.Device object at 0x7f41d4098688>
0.06576069355010987
0.006289639472961426
error gx2= 2.0638987e-07
error gy= 1.0912560261408778e-07
acceleration= 10.455399523742015
17.97926664352417 2.710083246231079
acceleration in solver= 6.634211944790991

Contact information

email: [e-mail]