Mark Kim
20328534b5
clean up warnings.
2018-12-13 14:49:38 -05:00
Mark Kim
9c86727225
3d zfp decompress filter.
2018-12-13 14:49:38 -05:00
Mark Kim
7655797321
3d zfp compressor filter.
2018-12-13 14:49:38 -05:00
Mark Kim
a6bcf8a0ab
something's not quite right.
2018-12-13 14:49:38 -05:00
Mark Kim
ce44a4d183
better tests.
2018-12-13 14:49:38 -05:00
Mark Kim
e033fe8eea
fix new 1d data generator.
2018-12-13 14:49:38 -05:00
Mark Kim
8a9bfcba1b
zfp 1d worklet.
2018-12-13 14:49:38 -05:00
Mark Kim
ece5913869
good to use the actual 1D compressor.
2018-12-13 14:49:38 -05:00
Mark Kim
aad7498387
1d compressor filter
2018-12-13 14:49:37 -05:00
Mark Kim
c08a0bf3f8
visual check of compression/decompression
2018-12-13 14:49:37 -05:00
Mark Kim
851c424a16
fix some bugs.
2018-12-13 14:49:37 -05:00
Mark Kim
cb8a5fd9c8
update 2d zfp filter to new 2d encode/decode worklets.
2018-12-13 14:49:37 -05:00
Mark Kim
95636ff2cd
decode/encode 2d.
2018-12-13 14:49:37 -05:00
Mark Kim
d06190fdc2
how will we pass the original type?
2018-12-13 14:49:37 -05:00
Mark Kim
b41b89312c
2d zfp compress/decompress and tests.
2018-12-13 14:49:37 -05:00
Mark Kim
7118139b93
add filter testing for zfp 1d.
2018-12-13 14:49:37 -05:00
Mark Kim
2026eba609
forgot stuff.
2018-12-13 14:49:37 -05:00
Mark Kim
3b99722f6e
Prototyping zfp 1d filter.
2018-12-13 14:49:37 -05:00
mclarsen
aa891c58a7
fixed decompression porting bugs
2018-12-13 14:49:37 -05:00
mclarsen
37cba9e99f
adding path for decompress
2018-12-13 14:49:37 -05:00
mclarsen
0c788b92a2
fixed error with integer types
2018-12-13 14:49:37 -05:00
mclarsen
584b36d64e
adding partial gather3
2018-12-13 14:49:37 -05:00
mclarsen
4650881f58
something
2018-12-13 14:49:37 -05:00
mclarsen
1020cb43e6
more missing files
2018-12-13 14:49:37 -05:00
mclarsen
cf3d24d1aa
more missing files
2018-12-13 14:49:37 -05:00
mclarsen
baa0ad77c7
adding missing file
2018-12-13 14:49:36 -05:00
mclarsen
bcd0150e9e
save point
2018-12-13 14:49:36 -05:00
mclarsen
1c45e48fac
compression working and benchmarked
2018-12-13 14:49:36 -05:00
mclarsen
c01e1b7349
debugging cuda failure
2018-12-13 14:49:36 -05:00
mclarsen
8cd7c5db32
fixed the gather
2018-12-13 14:49:36 -05:00
mclarsen
ea9bef19da
init zfp
2018-12-13 14:49:36 -05:00
Allison Vacanti
f3536e44f8
Merge topic '320-vtkm_init'
...
cdb1f5680 Add vtkm::cont::Initialize.
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Kenneth Moreland <kmorel@sandia.gov>
Acked-by: Haocheng LIU <haocheng.liu@kitware.com>
Merge-request: !1475
2018-12-13 12:01:57 -05:00
Allison Vacanti
cdb1f5680a
Add vtkm::cont::Initialize.
...
Also
- Renamed vtkm::cont::make_DeviceAdapterIdFromName to just overload
make_DeviceAdapterId.
- Refactored CMake logic for unit tests
- Since we're now querying the device tracker for the names, they
cannot be all caps.
- Updated usages of InitLogging to use Initialize instead.
- Added changelog.
2018-12-13 10:15:44 -05:00
Kenneth Moreland
6e87ac5628
Merge topic 'count-to-offset-use-device-id'
...
81a22ff64 Change ConvertNumComponentsToOffsets to use DeviceAdapterId
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Robert Maynard <robert.maynard@kitware.com>
Merge-request: !1479
2018-12-12 14:00:02 -05:00
Kenneth Moreland
81a22ff64f
Change ConvertNumComponentsToOffsets to use DeviceAdapterId
...
The vtkm::cont::ConvertNumComponentsToOffsets method was originally
created before vtkm::cont::DeviceAdapterId and the dynamic
vtkm::cont::Algorithm classes were created. Thus, it contained its own
version of TryExecute and did not support the direct selection of a
device at runtime. Change the function to use the new dynamic device
adapter id, which makes this more consistent with the rest of VTK-m and
cleans up the implementation quite a bit.
2018-12-12 08:18:47 -07:00
Robert Maynard
2ca0fecf96
Merge topic 'correct_is_trivially_for_intel'
...
499419006 UnitTestMoveConstructors will now compile with gcc 4.9 and Intel < 17.
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Allison Vacanti <allison.vacanti@kitware.com>
Merge-request: !1478
2018-12-11 10:15:19 -05:00
Robert Maynard
4994190064
UnitTestMoveConstructors will now compile with gcc 4.9 and Intel < 17.
2018-12-11 09:29:08 -05:00
Robert Maynard
83d5af6159
Merge topic 'improve_vtkm_cont_classes_to_have_noexcept_move_ops'
...
0da7830d9 Make sure important vtkm::cont classes have noexcept move ops.
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Kenneth Moreland <kmorel@sandia.gov>
Merge-request: !1477
2018-12-10 11:13:01 -05:00
Robert Maynard
0da7830d9a
Make sure important vtkm::cont classes have noexcept move ops.
...
This is important so they can be stored efficiently inside
containers such as std::vector that require strong move guarantees.
2018-12-10 10:33:18 -05:00
Robert Maynard
6520ca1ff9
Merge topic 'improve_array_handle_basic_serialization_perf'
...
fee6ef591 Improve serialization performance of ArrayHandle with BasicStorage
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Allison Vacanti <allison.vacanti@kitware.com>
Merge-request: !1474
2018-12-07 10:22:41 -05:00
Robert Maynard
dd0cbc3689
Merge topic 'disable_cuda_optimizations_for_UnitTestBoundingIntervalHierarchy'
...
a9c151188 Disable ptxas for CUDA 8 && 9 for UnitTestBoundingIntervalHierarchy.
Acked-by: Kitware Robot <kwrobot@kitware.com>
Merge-request: !1465
2018-12-07 08:28:04 -05:00
Allison Vacanti
629b290216
Merge topic 'dev_flags'
...
8baacca52 Link vtkm_developer_flags directly into vtkm libraries.
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Robert Maynard <robert.maynard@kitware.com>
Merge-request: !1473
2018-12-06 15:11:39 -05:00
Allison Vacanti
8baacca525
Link vtkm_developer_flags directly into vtkm libraries.
...
This should properly propogate the flags to superprojects.
2018-12-06 13:56:18 -05:00
Robert Maynard
fee6ef591b
Improve serialization performance of ArrayHandle with BasicStorage
2018-12-06 09:21:06 -05:00
Kenneth Moreland
1766f213ea
Merge topic 'dynamic-neighborhood-size'
...
82536092c Rename boundary classes
e008aff27 Change neighborhood parameter from layer to radius
7f627380b Fix Marching Cubes regression test
01a9e8541 Change WorkletPointNeighborhood to specify neighborhood at runtime
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Robert Maynard <robert.maynard@kitware.com>
Merge-request: !1457
2018-12-04 16:02:46 -05:00
Sujin Philip
4169ff7dc2
Merge topic 'implement-extents-v2'
...
6d81bc8b0 Implement extents support
Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Kenneth Moreland <kmorel@sandia.gov>
Merge-request: !1472
2018-12-04 09:40:32 -05:00
Sujin Philip
6d81bc8b06
Implement extents support
...
Adds a variable `GlobalPointIndexStart` to `CellSetStructured`.
Adding this to the cell-set, instead of the coordinate system, enables this
feature for different types of datasets like uniform grid, rectilinear, etc.,
with this one change.
The extents can be computed using `GlobalPointIndexStart` and `PointDimensions`.
2018-12-03 15:56:45 -05:00
Kenneth Moreland
82536092cc
Rename boundary classes
...
Did a bit of renaming of the support classes used for
WorkletPointNeighborhood. First, the OnBoundary tag is changed to
Boundary to match other tags and reflect some changes in the resulting
methods. Also moved the BoundaryState and Neighborhood classes from
vtkm::exec::arg to vtkm::exec to be more accessible. Finally, the
Neighborhood class name was changed to FieldNeighborhood to be more
specific on what role this class plays with neighborhood.
2018-12-03 12:45:23 -07:00
Kenneth Moreland
e008aff27b
Change neighborhood parameter from layer to radius
...
It was determined that the word "radius" more clearly described how the
size of the neighborhood is determined.
2018-12-03 12:45:23 -07:00
Kenneth Moreland
7f627380b6
Fix Marching Cubes regression test
...
Turns out there was a bug in the way Marching Cubes computed gradients
on a structured grid. A recent change to Boundary also had a similar bug
that ended in the same behavior. The recent changes to BoundaryState
fixed the problem, but caused the Marching Cubes test to fail because it
was comparing against bad gradient values.
This change fixes the array of gradient values. I'm pretty sure these
are correct since I generated them with ParaView (using a completely
separate implementation).
2018-12-03 12:45:23 -07:00