Merge branch 'main' into slava/hypersphere_improvements

CHANGELOG.md

@@ -1,5 +1,17 @@
 # CHANGELOG
 
+## v0.2.1 - 08.08.2024
+ Minor release with mostly cosmetic changes:
+* Add "If you have a question" section to README.md by @vabor112 in https://github.com/geometric-kernels/GeometricKernels/pull/131
+* Github cosmetics by @stoprightthere in https://github.com/geometric-kernels/GeometricKernels/pull/133
+* Replace all references to "gpflow" organization with "geometric-kernels" organization by @vabor112 in https://github.com/geometric-kernels/GeometricKernels/pull/134
+* Use fit_gpytorch_model or fit.fit_gpytorch_mll depening on the botorсh version by @vabor112 in https://github.com/geometric-kernels/GeometricKernels/pull/137
+* Add a missing type cast and fix a typo in kernels/karhunen_loeve.py by @vabor112 in https://github.com/geometric-kernels/GeometricKernels/pull/136
+* Minor documentation improvements by @vabor112 in https://github.com/geometric-kernels/GeometricKernels/pull/135
+* Add citation to the preprint of the GeometricKernels paper by @vabor112 in https://github.com/geometric-kernels/GeometricKernels/pull/138
+* Add citation file by @aterenin in https://github.com/geometric-kernels/GeometricKernels/pull/140
+* Fix dependencies (Version 0.2.1) by @stoprightthere in https://github.com/geometric-kernels/GeometricKernels/pull/143
+
 ## v0.2 - 21.04.2024
  New geometric kernel that *just works*, `kernels.MaternGeometricKernel`. Relies on *(hopefully)* sensible defaults we defined. Mostly by @stoprightthere.
 

CITATION.bib

@@ -0,0 +1,6 @@
+@article{mostowsky2024,
+      title = {The GeometricKernels Package: Heat and Matérn Kernels for Geometric Learning on Manifolds, Meshes, and Graphs},
+      author = {Peter Mostowsky and Vincent Dutordoir and Iskander Azangulov and Noémie Jaquier and Michael John Hutchinson and Aditya Ravuri and Leonel Rozo and Alexander Terenin and Viacheslav Borovitskiy},
+      year = {2024},
+      journal = {arXiv:2407.08086},
+}

CITATION.cff

@@ -0,0 +1,29 @@
+cff-version: 1.2.0
+message: "If you use this software, please cite it as below."
+title: "GeometricKernels"
+authors:
+- name: "GeometricKernels Contributors"
+preferred-citation:
+  type: "article"
+  title: "The GeometricKernels Package: Heat and Matérn Kernels for Geometric Learning on Manifolds, Meshes, and Graphs"
+  authors:
+  - family-names: "Mostowsky"
+    given-names: "Peter"
+  - family-names: "Dutordoir"
+    given-names: "Vincent"
+  - family-names: "Azangulov"
+    given-names: "Iskander"
+  - family-names: "Jaquier"
+    given-names: "Noémie"
+  - family-names: "Hutchinson"
+    given-names: "Michael John"
+  - family-names: "Ravuri"
+    given-names: "Aditya"
+  - family-names: "Rozo"
+    given-names: "Leonel"
+  - family-names: "Terenin"
+    given-names: "Alexander"
+  - family-names: "Borovitskiy"
+    given-names: "Viacheslav"
+  year: "2024"
+  journal: "arXiv:2407.08086"

geometric_kernels/feature_maps/probability_densities.py

@@ -13,7 +13,6 @@
 import lab as B
 import numpy as np
 from beartype.typing import Dict, List, Optional, Tuple
-from opt_einsum import contract as einsum
 from sympy import Poly, Product, symbols
 
 from geometric_kernels.lab_extras import (
@@ -85,7 +84,7 @@ def student_t_sample(
     shape_sqrt = B.chol(shape)
     dtype = dtype or dtype_double(key)
     key, z = B.randn(key, dtype, *size, n)
-    z = einsum("...i,ji->...j", z, shape_sqrt)
+    z = B.einsum("...i,ji->...j", z, shape_sqrt)
 
     key, g = B.randgamma(
         key,

geometric_kernels/kernels/feature_map.py

@@ -7,7 +7,6 @@
 import lab as B
 import numpy as np
 from beartype.typing import Dict, Optional
-from opt_einsum import contract as einsum
 
 from geometric_kernels.feature_maps import FeatureMap
 from geometric_kernels.kernels.base import BaseGeometricKernel
@@ -124,7 +123,7 @@ def K(
         else:
             features_X2 = features_X
 
-        feature_product = einsum("...no,...mo->...nm", features_X, features_X2)
+        feature_product = B.einsum("...no,...mo->...nm", features_X, features_X2)
         return feature_product
 
     def K_diag(self, params: Dict[str, B.Numeric], X: B.Numeric, **kwargs):

geometric_kernels/spaces/eigenfunctions.py

@@ -22,7 +22,6 @@
 
 import lab as B
 from beartype.typing import List, Optional
-from opt_einsum import contract as einsum
 
 from geometric_kernels.lab_extras import complex_like, is_complex, take_along_axis
 
@@ -106,7 +105,7 @@ def weighted_outerproduct(
         else:
             sum_phi_phi_for_level = B.cast(B.dtype(weights), sum_phi_phi_for_level)
 
-        return einsum("id,...nki->...nk", weights, sum_phi_phi_for_level)  # [N, N2]
+        return B.einsum("id,...nki->...nk", weights, sum_phi_phi_for_level)  # [N, N2]
 
     def weighted_outerproduct_diag(
         self, weights: B.Numeric, X: B.Numeric, **kwargs
@@ -132,7 +131,7 @@ def weighted_outerproduct_diag(
         else:
             phi_product_diag = B.cast(B.dtype(weights), phi_product_diag)
 
-        return einsum("id,ni->n", weights, phi_product_diag)  # [N,]
+        return B.einsum("id,ni->n", weights, phi_product_diag)  # [N,]
 
     @abc.abstractmethod
     def phi_product(
@@ -301,7 +300,7 @@ def phi_product(
             X2 = X
         Phi_X = self.__call__(X, **kwargs)  # [N, J]
         Phi_X2 = self.__call__(X2, **kwargs)  # [N2, J]
-        return einsum("nl,ml->nml", Phi_X, Phi_X2)  # [N, N2, J]
+        return B.einsum("nl,ml->nml", Phi_X, Phi_X2)  # [N, N2, J]
 
     def phi_product_diag(self, X: B.Numeric, **kwargs):
         Phi_X = self.__call__(X, **kwargs)  # [N, J]

geometric_kernels/spaces/hyperbolic.py

@@ -5,7 +5,6 @@
 import geomstats as gs
 import lab as B
 from beartype.typing import Optional
-from opt_einsum import contract as einsum
 
 from geometric_kernels.lab_extras import (
     complex_like,
@@ -134,7 +133,7 @@ def inner_product(self, vector_a, vector_b):
         p = 1
         diagonal = from_numpy(vector_a, [-1.0] * p + [1.0] * q)  # (n+1)
         diagonal = B.cast(B.dtype(vector_a), diagonal)
-        return einsum("...i,...i->...", diagonal * vector_a, vector_b)
+        return B.einsum("...i,...i->...", diagonal * vector_a, vector_b)
 
     def inv_harish_chandra(self, lam: B.Numeric) -> B.Numeric:
         lam = B.squeeze(lam, -1)

geometric_kernels/spaces/so.py

@@ -13,7 +13,6 @@
 import lab as B
 import numpy as np
 from beartype.typing import List, Tuple
-from opt_einsum import contract as einsum
 
 from geometric_kernels.lab_extras import dtype_double, from_numpy, qr, take_along_axis
 from geometric_kernels.spaces.eigenfunctions import Eigenfunctions
@@ -300,7 +299,7 @@ def random(self, key: B.RandomState, number: int):
             # explicit parametrization via the double cover SU(2) = S^3
             key, sphere_point = B.random.randn(key, dtype_double(key), number, 4)
             sphere_point /= B.reshape(
-                B.sqrt(einsum("ij,ij->i", sphere_point, sphere_point)), -1, 1
+                B.sqrt(B.einsum("ij,ij->i", sphere_point, sphere_point)), -1, 1
             )
 
             x, y, z, w = (B.reshape(sphere_point[..., i], -1, 1) for i in range(4))
@@ -317,7 +316,7 @@ def random(self, key: B.RandomState, number: int):
             # qr decomposition is not in the lab package, so numpy is used.
             key, h = B.random.randn(key, dtype_double(key), number, self.n, self.n)
             q, r = qr(h, mode="complete")
-            r_diag_sign = B.sign(einsum("...ii->...i", r))
+            r_diag_sign = B.sign(B.einsum("...ii->...i", r))
             q *= r_diag_sign[:, None]
             q_det_sign = B.sign(B.det(q))
             q[:, :, 0] *= q_det_sign[:, None]

geometric_kernels/spaces/su.py

@@ -12,7 +12,6 @@
 import lab as B
 import numpy as np
 from beartype.typing import List, Tuple
-from opt_einsum import contract as einsum
 
 from geometric_kernels.lab_extras import (
     complex_conj,
@@ -222,7 +221,7 @@ def random(self, key: B.RandomState, number: int):
             # explicit parametrization via the double cover SU(2) = S_3
             key, sphere_point = B.random.randn(key, dtype_double(key), number, 4)
             sphere_point /= B.reshape(
-                B.sqrt(einsum("ij,ij->i", sphere_point, sphere_point)), -1, 1
+                B.sqrt(B.einsum("ij,ij->i", sphere_point, sphere_point)), -1, 1
             )
             a = create_complex(sphere_point[..., 0], sphere_point[..., 1])
             b = create_complex(sphere_point[..., 2], sphere_point[..., 3])
@@ -235,7 +234,7 @@ def random(self, key: B.RandomState, number: int):
             key, imag = B.random.randn(key, dtype_double(key), number, self.n, self.n)
             h = create_complex(real, imag) / B.sqrt(2)
             q, r = qr(h, mode="complete")
-            r_diag = einsum("...ii->...i", r)
+            r_diag = B.einsum("...ii->...i", r)
             r_diag_inv_phase = complex_conj(r_diag / B.abs(r_diag))
             q *= r_diag_inv_phase[:, None]
             q_det = B.det(q)

pyproject.toml

@@ -23,18 +23,19 @@ keywords=[
 ]
 requires-python = ">=3.8"
 dependencies = [
+    "backends>=1.5.4",
+    "einops",
+    "geomstats",
     "numpy>=1.16",
-    "scipy>=1.3",
+    "opt-einsum",
     "plum-dispatch>=2.2.0",
-    "backends>=1.5.4",
     "potpourri3d",
     "robust_laplacian",
-    "opt-einsum",
-    "geomstats",
-    "einops",
+    "scipy>=1.3",
     "spherical-harmonics-basis",
+    "sympy~=1.13",
 ]
-version="0.2"
+version="0.2.1"
 
 [project.urls]
 Documentation = "https://geometric-kernels.github.io/"

tests/kernels/test_product.py

@@ -1,6 +1,5 @@
 import lab as B
 import numpy as np
-from opt_einsum import contract as einsum
 
 from geometric_kernels.kernels import MaternKarhunenLoeveKernel, ProductGeometricKernel
 from geometric_kernels.lab_extras.extras import from_numpy
@@ -43,7 +42,7 @@ def test_circle_product_eigenfunctions():
     weights = B.expand_dims(weights, -1)
     actual = B.to_numpy(eigenfunctions.weighted_outerproduct(weights, X, X))
 
-    expected = einsum("ni,mi,i->nm", Phi_X, Phi_X2, chained_weights)
+    expected = B.einsum("ni,mi,i->nm", Phi_X, Phi_X2, chained_weights)
     np.testing.assert_array_almost_equal(actual, expected)
 
 

tests/spaces/test_circle.py

@@ -3,7 +3,6 @@
 import pytest
 import tensorflow as tf
 import torch
-from opt_einsum import contract as einsum
 from plum import Tuple
 
 from geometric_kernels.kernels import MaternKarhunenLoeveKernel
@@ -89,7 +88,7 @@ def test_weighted_outerproduct_with_addition_theorem(
 
     Phi_X = eigenfunctions(inputs)
     Phi_X2 = eigenfunctions(inputs2)
-    expected = einsum("ni,ki,i->nk", Phi_X, Phi_X2, chained_weights)
+    expected = B.einsum("ni,ki,i->nk", Phi_X, Phi_X2, chained_weights)
     np.testing.assert_array_almost_equal(actual, expected)
 
 
@@ -124,7 +123,7 @@ def test_weighted_outerproduct_diag_with_addition_theorem(
     actual = eigenfunctions.weighted_outerproduct_diag(weights, inputs)
 
     Phi_X = eigenfunctions(inputs)
-    expected = einsum("ni,i->n", Phi_X**2, chained_weights)
+    expected = B.einsum("ni,i->n", Phi_X**2, chained_weights)
     np.testing.assert_array_almost_equal(B.to_numpy(actual), B.to_numpy(expected))
 
 

tests/spaces/test_hypersphere.py

@@ -1,7 +1,6 @@
 import lab as B
 import numpy as np
 import pytest
-from opt_einsum import contract as einsum
 from plum import Tuple
 
 from geometric_kernels.spaces.hypersphere import SphericalHarmonics
@@ -77,7 +76,7 @@ def test_weighted_outerproduct_with_addition_theorem(
 
     Phi_X = eigenfunctions(inputs)
     Phi_X2 = eigenfunctions(inputs2)
-    expected = einsum("ni,ki,i->nk", Phi_X, Phi_X2, chained_weights)
+    expected = B.einsum("ni,ki,i->nk", Phi_X, Phi_X2, chained_weights)
     np.testing.assert_array_almost_equal(actual, expected)
 
 
@@ -112,5 +111,5 @@ def test_weighted_outerproduct_diag_with_addition_theorem(
     actual = eigenfunctions.weighted_outerproduct_diag(weights, inputs)
 
     Phi_X = eigenfunctions(inputs)
-    expected = einsum("ni,i->n", Phi_X**2, chained_weights)
+    expected = B.einsum("ni,i->n", Phi_X**2, chained_weights)
     np.testing.assert_array_almost_equal(B.to_numpy(actual), B.to_numpy(expected))

tests/spaces/test_lie_groups.py

@@ -4,7 +4,6 @@
 import numpy as np
 import pytest
 from numpy.testing import assert_allclose
-from opt_einsum import contract as einsum
 
 from geometric_kernels.feature_maps import RandomPhaseFeatureMapCompact
 from geometric_kernels.kernels import MaternKarhunenLoeveKernel
@@ -127,6 +126,6 @@ def test_feature_map(group_and_eigf):
 
     K_xx = (kernel.K(param, x, x)).real
     key, embed_x = feature_map(x, param, key=key, normalize=True)
-    F_xx = (einsum("ni,mi-> nm", embed_x, embed_x.conj())).real
+    F_xx = (B.einsum("ni,mi-> nm", embed_x, embed_x.conj())).real
 
     assert_allclose(K_xx, F_xx, atol=5e-2)