Merge pull request #302 from JuliaRobotics/maint/mani_v0.10

Update to Manifolds v0.10 SE(n, vectors=HybridTangentRepresentation())

Project.toml

@@ -1,7 +1,7 @@
 name = "ApproxManifoldProducts"
 uuid = "9bbbb610-88a1-53cd-9763-118ce10c1f89"
 keywords = ["MM-iSAMv2", "SLAM", "inference", "sum-product", "belief-propagation", "nonparametric", "manifolds", "functional"]
-version = "0.8.5"
+version = "0.8.6"
 
 [deps]
 Base64 = "2a0f44e3-6c83-55bd-87e4-b1978d98bd5f"
@@ -42,7 +42,7 @@ CoordinateTransformations = "0.5, 0.6"
 Distributions = "0.25"
 DocStringExtensions = "0.7, 0.8, 0.9"
 KernelDensityEstimate = "0.5.10"
-Manifolds = "0.9"
+Manifolds = "0.10"
 ManifoldsBase = "0.14, 0.15"
 NLsolve = "3, 4"
 Optim = "1"

examples/AMP41_dev.jl

@@ -29,7 +29,7 @@ end
 
 ##
 
-M = SpecialEuclidean(2)
+M = SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 N = 128
 
 

examples/dev_BallTreeDensity.jl

@@ -3,7 +3,7 @@
 
 using StaticArrays, Manifolds, NearestNeighbors, Distances
 
-M = SpecialEuclidean(2)
+M = SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 N = 100
 # convert point to coordinates
 function coords(p)
@@ -166,7 +166,7 @@ distance(mr.manifold, SVector(1,2),SVector(2,3))
 ##
 
 
-M_se2 = SpecialEuclidean(2)
+M_se2 = SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 
 
 sI = SMatrix{3,3,Float64}(diagm(ones(3)))

src/entities/ManifoldDefinitions.jl

@@ -37,8 +37,8 @@ const Euclid3 = TranslationGroup(3)
 const Euclid4 = TranslationGroup(4)
 
 # TODO if not easy simplification exists, then just deprecate this
-const SE2_Manifold = SpecialEuclidean(2)
-const SE3_Manifold = SpecialEuclidean(3)
+const SE2_Manifold = SpecialEuclidean(2; vectors=HybridTangentRepresentation())
+const SE3_Manifold = SpecialEuclidean(3; vectors=HybridTangentRepresentation())
 
 
 

src/services/ManifoldPartials.jl

@@ -126,7 +126,7 @@ using Manifolds
 using ApproxManifoldProducts
 
 # a familiar manifold of translation and rotation in 2D
-M = SpecialEuclidean(2)
+M = SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 
 
 # make a new partial of only the translation components

src/services/ManifoldsOverloads.jl

@@ -11,8 +11,8 @@ LieGroupManifoldsPirate = Union{
   typeof(TranslationGroup(6)),
   typeof(SpecialOrthogonal(2)), 
   typeof(SpecialOrthogonal(3)), 
-  typeof(SpecialEuclidean(2)), 
-  typeof(SpecialEuclidean(3))
+  typeof(SpecialEuclidean(2; vectors=HybridTangentRepresentation())), 
+  typeof(SpecialEuclidean(3; vectors=HybridTangentRepresentation()))
 }
 
 ## ===================================== BASIS PIRATES =====================================
@@ -46,15 +46,15 @@ get_basis_affine(
 )
 
 get_basis_affine(
-  ::typeof(SpecialEuclidean(2))
+  ::typeof(SpecialEuclidean(2; vectors=HybridTangentRepresentation()))
 ) = tuple(
   SA[0 -0 1; 0 0 0; 0 0 0.0],
   SA[0 -0 0; 0 0 1; 0 0 0.0],
   SA[0 -1 0; 1 0 0; 0 0 0.0],
 )
 
 get_basis_affine(
-  ::typeof(SpecialEuclidean(3))
+  ::typeof(SpecialEuclidean(3; vectors=HybridTangentRepresentation()))
 ) = tuple(
   SA[0 -0 0 1; 0 0 -0 0; -0 0 0 0; 0 0 0 0.0],
   SA[0 -0 0 0; 0 0 -0 1; -0 0 0 0; 0 0 0 0.0],
@@ -141,7 +141,7 @@ parallel_transport_best(
 
 
 _makeaffine(::AbstractManifold, X) = X
-_makeaffine(M::Union{typeof(SpecialEuclidean(2)),typeof(SpecialEuclidean(3))}, X::ArrayPartition) = screw_matrix(M,X)
+_makeaffine(M::Union{typeof(SpecialEuclidean(2; vectors=HybridTangentRepresentation())),typeof(SpecialEuclidean(3; vectors=HybridTangentRepresentation()))}, X::ArrayPartition) = screw_matrix(M,X)
 
 # assumes inputs are Lie algebra tangent vectors represented in matrix form
 ad(
@@ -197,7 +197,7 @@ end
 
 
 function Ad(
-  M::Union{typeof(SpecialEuclidean(2)), typeof(SpecialEuclidean(3))}, 
+  M::Union{typeof(SpecialEuclidean(2; vectors=HybridTangentRepresentation())), typeof(SpecialEuclidean(3; vectors=HybridTangentRepresentation()))}, 
   p, 
   X::ArrayPartition;
   use_upstream::Bool = _UPSTREAM_MANIFOLDS_ADJOINT_ACTION # explict to support R&D
@@ -311,7 +311,7 @@ Ad(
 # d is a Lie algebra element (tangent vector) providing the direction of transport
 # X is the tangent vector to be transported 
 function ad(
-  M::typeof(SpecialEuclidean(3)), 
+  M::typeof(SpecialEuclidean(3; vectors=HybridTangentRepresentation())), 
   d::ArrayPartition, 
   X::ArrayPartition
 )
@@ -332,7 +332,7 @@ end
 
 
 function ad(
-  ::typeof(SpecialEuclidean(2)), 
+  ::typeof(SpecialEuclidean(2; vectors=HybridTangentRepresentation())), 
   d::ArrayPartition, 
 )
   Vx = SA[d.x[1][2]; -d.x[1][1]]
@@ -344,7 +344,7 @@ function ad(
 end
 
 function ad(
-  ::typeof(SpecialEuclidean(3)), 
+  ::typeof(SpecialEuclidean(3; vectors=HybridTangentRepresentation())), 
   d::ArrayPartition, 
 )
   Vx = skew(d.x[1])
@@ -357,7 +357,7 @@ end
 
 
 function Ad(
-  ::typeof(SpecialEuclidean(2)), 
+  ::typeof(SpecialEuclidean(2; vectors=HybridTangentRepresentation())), 
   p
 )
   t = p.x[1]
@@ -369,7 +369,7 @@ function Ad(
 end
 
 function Ad(
-  ::typeof(SpecialEuclidean(3)), 
+  ::typeof(SpecialEuclidean(3; vectors=HybridTangentRepresentation())), 
   p
 )
   t = p.x[1]

test/basic_se3.jl

@@ -6,7 +6,7 @@ using Test
 
 ##
 
-@testset "Test isapprox function on basic SpecialEuclidean(3)" begin
+@testset "Test isapprox function on basic SpecialEuclidean(3; vectors=HybridTangentRepresentation())" begin
 
 ##
 
@@ -17,7 +17,7 @@ using Test
 # c_[1,:] .+= 50
 # c = kde!(c_)
 
-M = SpecialEuclidean(3)
+M = SpecialEuclidean(3; vectors=HybridTangentRepresentation())
 u0 = ArrayPartition(zeros(3),[1 0 0; 0 1 0; 0 0 1.0])
 ϵ = identity_element(M, u0)
 N = 50

test/manellic/testManellicTree.jl

@@ -309,7 +309,7 @@ ker = AMP.MvNormalKernel([0], [2.0;;])
 )
 
 ##
-M = SpecialEuclidean(2)
+M = SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 ε = identity_element(M)
 Xc = [10, 20, 0.1]
 p = exp(M, ε, hat(M, ε, Xc))
@@ -424,7 +424,7 @@ ys_pdf = pdf(dis, ps)
 # lines!(first.(ps), ys_amp)
 
 
-M = SpecialEuclidean(2)
+M = SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 ε = identity_element(M)
 dis = MvNormal([10,20,0.1], diagm([0.5,2.0,0.1].^2)) 
 Cpts = [rand(dis) for _ in 1:128]
@@ -447,9 +447,9 @@ end
 
 
 ## ========================================================================================
-@testset "Test Product the brute force way, SpecialEuclidean(2)" begin
+@testset "Test Product the brute force way, SpecialEuclidean(2; vectors=HybridTangentRepresentation())" begin
 
-M = SpecialEuclidean(2)
+M = SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 ε = identity_element(M)
 
 Xc_p = [10, 20, 0.1]
@@ -610,10 +610,10 @@ maj_ang = (angle(Complex(evv.vectors[:,maj_idx]...)) + 2pi) % pi
 end
 
 
-@testset "Rotated covariance product major axis checks, SpecialEuclidean(2)" begin
+@testset "Rotated covariance product major axis checks, SpecialEuclidean(2; vectors=HybridTangentRepresentation())" begin
 ##
 
-M = SpecialEuclidean(2)
+M = SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 ε = identity_element(M)
 
 Xc_p = [0, 0, 0.0]
@@ -915,10 +915,10 @@ end
 
 
 
-@testset "Multidimensional LOOCV bandwidth optimization, SpecialEuclidean(2)" begin
+@testset "Multidimensional LOOCV bandwidth optimization, SpecialEuclidean(2; vectors=HybridTangentRepresentation())" begin
 ##
 
-M = SpecialEuclidean(2)
+M = SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 pts = [ArrayPartition(randn(2),Rot_.RotMatrix{2}(0.1*randn()).mat) for _ in 1:64]
 
 bw = [1.0; 1.0; 0.3]
@@ -954,10 +954,10 @@ end
 
 
 
-@testset "Multidimensional LOOCV bandwidth optimization, SpecialEuclidean(3)" begin
+@testset "Multidimensional LOOCV bandwidth optimization, SpecialEuclidean(3; vectors=HybridTangentRepresentation())" begin
 ##
 
-M = SpecialEuclidean(3)
+M = SpecialEuclidean(3; vectors=HybridTangentRepresentation())
 pts = [ArrayPartition(SA[randn(3)...;],SMatrix{3,3,Float64}(collect(Rot_.RotXYZ(0.1*randn(3)...)))) for _ in 1:64]
 
 bw = SA[1.0; 1.0; 1.0; 0.3; 0.3; 0.3]

test/testBasicManiProduct.jl

@@ -36,7 +36,7 @@ pts = AMP._pointsToMatrixCoords(P12.manifold, pts_)
 
 ##
 
-M = SpecialEuclidean(2)
+M = SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 u0 = ArrayPartition(zeros(2),[1 0; 0 1.0])
 ϵ = identity_element(M, u0)
 

test/testLieFundamentals.jl

@@ -175,7 +175,7 @@ end
 @testset "SE(2) Vector transport (w curvature) via Jacobians and Lie adjoints" begin
 ##
 
-M = SpecialEuclidean(2)
+M = SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 p = Identity(M)
 Xc = 0.5*randn(3)
 X = hat(M, p, Xc)   # random algebra element
@@ -208,7 +208,7 @@ end
 ##
 
 
-M = SpecialEuclidean(3)
+M = SpecialEuclidean(3; vectors=HybridTangentRepresentation())
 p = Identity(M)
 Xc = 0.5*randn(6)
 X = hat(M, p, Xc)   # random algebra element

test/testMKDStats.jl

@@ -7,7 +7,7 @@ using Manifolds
 @testset "Test basic MKD statistics" begin
 ##
 
-M = SpecialEuclidean(2)
+M = SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 u0 = ArrayPartition(zeros(2),[1 0; 0 1.0])
 ϵ = identity_element(M, u0)
 

test/testManiProductBigSmall.jl

@@ -11,7 +11,7 @@ using TensorCast
 
 ##
 
-M = SpecialEuclidean(2)
+M = SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 N = 100
 u0 = ArrayPartition([0;0.0],[1 0; 0 1.0])
 ϵ = identity_element(M, u0)

test/testManifoldConventions.jl

@@ -10,7 +10,7 @@ import Rotations as _Rot
 @testset "test Manifolds.jl conventions" begin
 ##
 
-M = SpecialEuclidean(2)
+M = SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 e0 = identity_element(M)
 
 # body to body next

test/testManifoldPartial.jl

@@ -57,13 +57,13 @@ M = Manifolds.Rotations(2)
 end
 
 
-@testset "test getManifoldPartial on SpecialEuclidean(2)" begin
+@testset "test getManifoldPartial on SpecialEuclidean(2; vectors=HybridTangentRepresentation())" begin
 
 ##
 
-M = SpecialEuclidean(2)
+M = SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 
-@test getManifoldPartial(M, [1;2;3])[1] == SpecialEuclidean(2)
+@test getManifoldPartial(M, [1;2;3])[1] == SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 
 @test getManifoldPartial(M, [1;])[1] == TranslationGroup(1)
 @test getManifoldPartial(M, [2;])[1] == TranslationGroup(1)
@@ -76,7 +76,7 @@ M = SpecialEuclidean(2)
 
 repr = ArrayPartition([0.0; 0], [1 0; 0 1.0])
 
-@test getManifoldPartial(M, [1;2;3], repr)[1] == SpecialEuclidean(2)
+@test getManifoldPartial(M, [1;2;3], repr)[1] == SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 @test getManifoldPartial(M, [1;2;3], repr)[2] == repr
 
 @test getManifoldPartial(M, [1;], repr)[1] == TranslationGroup(1)
@@ -117,7 +117,7 @@ end
 ##
 
 N = 100
-M = SpecialEuclidean(2)
+M = SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 u0 = ArrayPartition([0.0; 0], [1 0; 0 1.0])
 
 pts = [exp(M, u0, hat(M, u0, [10 .+ randn(2);randn()])) for i in 1:N]

test/testPartialProductSE2.jl

@@ -1,4 +1,4 @@
-# test on partial products with SpecialEuclidean(2)
+# test on partial products with SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 
 using Manifolds
 using ApproxManifoldProducts
@@ -7,7 +7,7 @@ using BSON
 
 ##
 
-@testset "partial product with a SpecialEuclidean(2)" begin
+@testset "partial product with a SpecialEuclidean(2; vectors=HybridTangentRepresentation())" begin
 ##
 
 datafile = joinpath(@__DIR__, "testdata", "partialtest.bson")
@@ -22,10 +22,10 @@ randN = Float64[]
 len = length(pts1)
 
 # define test manifold
-M = SpecialEuclidean(2)
+M = SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 e0 = ArrayPartition([0.0;0.0], [1 0; 0 1.0]) # identity_element(M)
 
-# p1 full SpecialEuclidean(2)
+# p1 full SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 p1 = manikde!(M, pts1)
 p1_ = marginal(p1,[1;2])
 

test/testSymmetry.jl

@@ -19,7 +19,7 @@ a,b = _Rot.RotMatrix(randn()), _Rot.RotMatrix(randn())
 @test isapprox( distance(M, a, b), distance(M, b, a), atol = 1e-5)
 
 
-M = SpecialEuclidean(2)
+M = SpecialEuclidean(2; vectors=HybridTangentRepresentation())
 a = ArrayPartition(randn(2),_Rot.RotMatrix(randn()))
 b = ArrayPartition(randn(2),_Rot.RotMatrix(randn()))
 @test isapprox( distance(M, a, b), distance(M, b, a), atol = 1e-5)