Changelog.md 2.62 KB
Newer Older
Boris Bonev's avatar
Boris Bonev committed
1
2
3
4
# Changelog

## Versioning

Boris Bonev's avatar
Boris Bonev committed
5
6
7
8
9
### v0.7.0

* CUDA-accelerated DISCO convolutions
* Updated DISCO convolutions to support even number of collocation points across the diameter
* Distributed DISCO convolutions
10
* Fused quadrature into multiplication with the Psi tensor to lower memory footprint
Boris Bonev's avatar
Boris Bonev committed
11
12
13
* Removed DISCO convolution in the plane to focus on the sphere
* Updated unit tests which now include tests for the distributed convolutions

Boris Bonev's avatar
Boris Bonev committed
14
15
### v0.6.5

Boris Bonev's avatar
Boris Bonev committed
16
17
18
19
20
* Discrete-continuous (DISCO) convolutions on the sphere and in two dimensions
* DISCO supports isotropic and anisotropic kernel functions parameterized as hat functions
* Supports regular and transpose convolutions
* Accelerated spherical DISCO convolutions on GPU via Triton implementation
* Unittests for DISCO convolutions in `tests/test_convolution.py`
Boris Bonev's avatar
Boris Bonev committed
21

22
### v0.6.4
Boris Bonev's avatar
Boris Bonev committed
23
24
25
26
27

* Reworking distributed to allow for uneven split tensors, effectively removing the necessity of padding the transformed tensors
* Distributed SHT tests are now using unittest. Test extended to vector SHT versions
* Tests are defined in `torch_harmonics/distributed/distributed_tests.py`
* Base pytorch container version bumped up to 23.11 in Dockerfile
28

Boris Bonev's avatar
Boris Bonev committed
29
30
31
### v0.6.3

* Adding gradient check in unit tests
32
* Temporary work-around for NCCL contiguous issues with distributed SHT
33
* Refactored examples and documentation
Boris Bonev's avatar
Boris Bonev committed
34
* Updated SFNO example
Boris Bonev's avatar
Boris Bonev committed
35

36
37
38
### v0.6.2

* Adding github CI
Boris Bonev's avatar
Boris Bonev committed
39
40
* Changed SHT modules to convert dtype dynamically when computing the SHT/ISHT
* Bugfixes to fix importing examples
41

Boris Bonev's avatar
Boris Bonev committed
42
43
### v0.6.1

Boris Bonev's avatar
Boris Bonev committed
44
* Minor bugfixes to export SFNO code
Boris Bonev's avatar
Boris Bonev committed
45
46
* Readme should now render correctly in PyPI

Boris Bonev's avatar
Boris Bonev committed
47
48
49
### v0.6.0

* Added SFNO example
50
* Added Shallow Water Equations Dataset for SFNO training
Boris Bonev's avatar
Boris Bonev committed
51
* Cleanup of the repository and added PyPI
52
* Updated Readme
Boris Bonev's avatar
Boris Bonev committed
53
54

### v0.5.0
Boris Bonev's avatar
Boris Bonev committed
55
56
57
58

* Reworked distributed SHT
* Module for sampling Gaussian Random Fields on the sphere

Boris Bonev's avatar
Boris Bonev committed
59
### v0.4.0
Boris Bonev's avatar
Boris Bonev committed
60
61
62
63
64

* Computation of associated Legendre polynomials
    * changed algorithm to compute the associated Legendre polynomials for improved stability
* Improved Readme

Boris Bonev's avatar
Boris Bonev committed
65
### v0.3.0
Boris Bonev's avatar
Boris Bonev committed
66
67
68
69
70
71
72
73
74
75
76
77
78
79

* Vector Spherical Harmonic Transforms
    * projects vector-valued fields onto the vector Spherical Harmonics
    * supports computation of div and curl on the sphere
* New quadrature rules
    * Clenshaw-Curtis quadrature rule
    * Fejér quadrature rule
    * Legendre-Gauss-Lobatto quadrature
* New notebooks
    * complete with differentiable Shallow Water Solver
    * notebook on quadrature and interpolation
* Unit tests
* Refactor of the API

Boris Bonev's avatar
Boris Bonev committed
80
### v0.2.0
Boris Bonev's avatar
Boris Bonev committed
81
82
83
84
85

* Renaming from torch_sht to torch_harmonics
* Adding distributed SHT support
* New logo

Boris Bonev's avatar
Boris Bonev committed
86
### v0.1.0
Boris Bonev's avatar
Boris Bonev committed
87
88

* Single GPU forward and backward transform
89
* Minimal code example and notebook