https://github.com/jyhmiinlin/pynufft

sort by:
Revision Author Date Message Commit Date
f6dc5b7 bump 13 February 2020, 03:39:17 UTC
4e08119 bump 13 February 2020, 03:37:14 UTC
3bea093 bump 13 February 2020, 03:35:55 UTC
c9cd3de README 03 February 2020, 12:49:12 UTC
c735523 edit the README 25 January 2020, 04:09:31 UTC
45454c7 Move to 2020.0.0. 13 January 2020, 19:47:41 UTC
f218ae2 fix batch=1. This can cause error in Riekna fft. 10 January 2020, 11:09:25 UTC
3a4ab17 Change version to 2019.2.3rc4 05 January 2020, 13:08:12 UTC
9a5cf9c Add test_init_device.py 05 January 2020, 12:49:42 UTC
76bc4df solve a missing import in legacy mode. Still, there is no support for batch mode in legacy mode. 26 December 2019, 19:37:43 UTC
4346979 add legacy code, which is fast for adjoint and cg 26 December 2019, 19:31:49 UTC
10dd4a3 Move linalg/nufft to a seperate module 26 December 2019, 11:28:19 UTC
4a11f62 add solve_device.py for nufft.NUFFT() when a device object is initialized 25 December 2019, 14:29:52 UTC
1dddeea merge subroutines to the NUFFT superclass. 23 December 2019, 15:11:19 UTC
4e15f0d Add virtualmethod plan() in nufft to switch between NUFFT_cpu.plan() and NUFFT_hsa.plan() Now NUFFT_cpu.plan() and NUFFT_hsa.plan() call their _plan() methods to avoid circular calling the function. 22 December 2019, 04:14:42 UTC
39cae6b Add helper.device_list() to list all accelerators. Add a class NUFFT which takes device_list()[num] as the input. 21 December 2019, 20:31:03 UTC
2f1a07f Some wording. Move to 2019.2.3rc2 06 December 2019, 08:19:21 UTC
abdc31d Fix test_installation() 11 November 2019, 07:59:36 UTC
02714f3 test_installation 11 November 2019, 07:58:45 UTC
72e5b19 Merge branch 'master' into release-2019.2.3 11 November 2019, 07:55:21 UTC
2f2e20a Correct Some Unforgivable Typos. 11 November 2019, 07:29:57 UTC
72e74d9 Merge branch 'release-2019.2.3' 09 November 2019, 19:34:28 UTC
926fd41 Fix typos in doc/source. 09 November 2019, 19:33:21 UTC
905f5bd remove unnecessary comments 03 November 2019, 10:21:03 UTC
47a7327 change the version to 2019.2.3rc1 03 November 2019, 09:05:52 UTC
bbc2471 add some example files 03 November 2019, 08:21:01 UTC
8f18289 change the version in setup.py to 2019.2.3 03 November 2019, 08:19:33 UTC
ce2b071 update contact information 03 November 2019, 08:11:53 UTC
3412d6b add nudft_cpu.py 22 September 2019, 11:25:02 UTC
964cbbc Change some documentations in nufft_cpu nufft_hsa 22 September 2019, 11:24:21 UTC
52cc97d Version 2019.2.2 20 September 2019, 14:01:14 UTC
2b0c353 modify many rst files 20 September 2019, 13:17:43 UTC
f11e733 copy some develop files to here 20 September 2019, 07:46:40 UTC
c6c82f8 bump 07 June 2019, 12:22:22 UTC
65b50ee modify the lsmr in solve_cpu 07 June 2019, 08:52:28 UTC
f17d6b9 bump 07 June 2019, 08:33:02 UTC
8cf0d6d scipy 1.3.0. removes scipy.misc.imresize. Change testing files. 07 June 2019, 08:25:23 UTC
0acee3b bump 21 May 2019, 09:13:49 UTC
6c2489a Move to v 2019.2.0 21 May 2019, 09:12:16 UTC
e9261fd update README 19 May 2019, 14:15:59 UTC
e0e658b bump to 2019.1.2 19 May 2019, 13:58:21 UTC
e8a864f BUGFIX: fix the loss of accuracy in cSpmvh() The original pELL_spmvh_mCoil was too complex and the group/local sizes have undetermined run-time behavior. Here I replace the function with a more simple logic. The global size is non-Cartesian numbers (M) 19 May 2019, 12:38:56 UTC
b2379df Fix the RuntimeWarning: overflow encountered in uint_scalars self.pELL['prodJd'] * self.batch) in the line 522 nufft_hsa.py 10 May 2019, 17:52:37 UTC
8919840 Cleaning some part of the code and removing useless prints (#18) * Initial commit of documentation fixing * [NUFFT_HSA]: Adding verbosity level and warning instead of prints * [NUFFT_hsa]: now pep8 compliant * [Helper]: Adding the verbosity * [NUFFT_hsa]: Adding the verbosity on the helper * [NUFFT_hsa]: pycodestyle issues * [nufft_cpu]: pycodestyle compatibility 10 May 2019, 09:14:34 UTC
6742ec5 update README.md 18 January 2019, 13:10:03 UTC
5321dd8 Rmove benchmark/benchmark_3D_mCoil.py 18 January 2019, 12:46:38 UTC
30c6fb3 use cuda's atomicAdd instead 17 January 2019, 10:42:53 UTC
bcc980d git status 17 January 2019, 10:22:58 UTC
8d45a80 Merge branch 'master' into develop 17 January 2019, 10:19:23 UTC
b42a23b update version to 2019.1.1rc1 17 January 2019, 10:19:06 UTC
2fbcb83 Merge branch 'master' into develop 17 January 2019, 10:17:33 UTC
57f59d8 bump 17 January 2019, 10:17:27 UTC
7636de5 bump 17 January 2019, 10:17:12 UTC
6650fbe Merge branch 'develop' 17 January 2019, 10:16:42 UTC
2361b3d remove unrelated file 17 January 2019, 10:13:57 UTC
cc42997 bump 17 January 2019, 10:08:08 UTC
3c979a5 bump, add some note for re_subroutine 16 January 2019, 10:19:41 UTC
f5c42e3 Update README.md, precompute with double complex in plan(). 15 January 2019, 21:31:43 UTC
0ec1b65 Add doc/source/manu/mcoil.rst 15 January 2019, 09:00:34 UTC
c1e2945 Change version to 2019.1.1 13 January 2019, 22:02:05 UTC
818df00 bump 13 January 2019, 21:59:49 UTC
8171512 Refactor the re_subroutine. Move the spread codes to a single file. 13 January 2019, 21:55:14 UTC
7e523c2 restructure the doc 13 January 2019, 12:30:59 UTC
b89ff01 Move all the scaling factors to radix-tensor and the final division in solve_cpu.cg. 13 January 2019, 12:28:39 UTC
3ce6eaa add benchmark/benchmark_3D_single_batch.py which tests the forward_single() and adjoint_single() methods fixed NUFFT_cpu.forward_single() and NUFFT_cpu.adjoint_single() 12 January 2019, 19:33:34 UTC
e2b2efb add reset_sense() in NUFFT_cpu class. 12 January 2019, 14:49:27 UTC
27a9613 Fixing the error of initial values of self.volume["gpu_coil_profile"].fill(1). 12 January 2019, 14:35:21 UTC
6e128e7 Fixed typo in NUFFT_hsa.x2s() 12 January 2019, 06:56:53 UTC
d06e8e1 Fix batched 'cg' in solve_hsa.py. Now 'cg' in solve_cpu and solve_hsa provide the same output. 09 January 2019, 20:59:17 UTC
a9b8dfb Make sure nufft_hsa pass the tests of tests/test_opencl_2D_multicoil.py, tests/test_opencl_2D.py 09 January 2019, 18:40:10 UTC
e0d4b73 Correct the split-radix function when radix is not a prime number of dimensions. Update the Windows doc 09 January 2019, 16:34:13 UTC
cf7ae7e Add tests/test_2D_cpu_batch.py. Tested for NUFFT_cpu. 07 January 2019, 22:12:40 UTC
7fdccd8 Merge branch 'develop' 07 January 2019, 14:46:46 UTC
a7586ff Only load test_init() in tests folder 07 January 2019, 14:29:18 UTC
1283925 Change MANIFEST.in 07 January 2019, 14:05:11 UTC
d4ca1a6 Move to Version 0.4.0-rc1 07 January 2019, 13:59:11 UTC
d7a08fe Update forward, forward_single, adjoint, adjoint_single, selfadjoint, selfadjoint_single in NUFFT_hsa Add set_sense metho in NUFFT_hsa 07 January 2019, 13:51:40 UTC
c9c8eda Major Update to 0.3.4.1. Add many files and documents. Add windows 10 installation. 05 January 2019, 21:04:58 UTC
2b88237 Tested under Windows 10. Change all uint to unsigned int for MSVC compatibility. Many updates. 05 January 2019, 20:59:24 UTC
6676d9b Add self.thr._context.push() to accommodate for cuda_api(). The reason is because multiple cuda contexts are stacked and only the current active context can be used. To call a method, the thr._context of the object being called must be pushed up to the top of the stack. Note: push() is not required for opencl interface. 15 December 2018, 20:58:18 UTC
61ea948 Add cHadamard.py and cMultiplyConjVecInplace.py into the git repository 14 December 2018, 16:07:40 UTC
5ccdcce Major update for GPU code. TO BE TESTED!! NOT for production! Move NUFFT_hsa class to NUFFT_hsa_legacy. Previous NUFFT_memsave class will replace the orginal NUFFT_hsa. Move CSR, CSRH, and pELL matrix contents to dictionaries in self.csr, self.csrH, and self.pELL. Move roll-off (scaling factor) and preindexed NdGPUorder, KdGPUorder to self.volume Add release() method which is called before delete the object. 14 December 2018, 16:03:22 UTC
47a9810 change wording in README 11 December 2018, 21:05:18 UTC
fcfecc5 Update README 11 December 2018, 21:00:40 UTC
eb26048 update to 0.3.3.12 11 December 2018, 20:20:00 UTC
04a93d7 Update the Pypi 11 December 2018, 20:19:27 UTC
9c17da0 bump 11 December 2018, 18:16:46 UTC
b6c983f bump 09 December 2018, 19:54:57 UTC
42243b7 bump 04 December 2018, 13:54:21 UTC
0144bef bump 04 December 2018, 13:52:22 UTC
6f445fa bump 03 December 2018, 08:48:54 UTC
c7198a2 bump 03 December 2018, 08:31:03 UTC
b7335b2 Bump 03 December 2018, 08:09:42 UTC
feba94b Bump 03 December 2018, 07:55:05 UTC
005d3c4 bump 03 December 2018, 07:36:44 UTC
7fc10d0 bump 27 November 2018, 14:04:19 UTC
03bc395 update README.md 09 November 2018, 14:23:40 UTC
6a2d4e5 update README.md for 3.3.10 09 November 2018, 10:58:29 UTC
a4b1674 Remove the keyword async for compatibility reasons because Reikna has changed the keyword to async_. 09 November 2018, 10:55:35 UTC
e3c33d1 change version to 3.3.11 09 November 2018, 10:37:18 UTC
back to top