func TestNormProbs(t *testing.T) {
dist1, ok := NewNormal([]float64{0, 0}, mat64.NewSymDense(2, []float64{1, 0, 0, 1}), nil)
if !ok {
t.Errorf("bad test")
}
dist2, ok := NewNormal([]float64{6, 7}, mat64.NewSymDense(2, []float64{8, 2, 0, 4}), nil)
if !ok {
t.Errorf("bad test")
}
testProbability(t, []probCase{
{
dist: dist1,
loc: []float64{0, 0},
logProb: -1.837877066409345,
},
{
dist: dist2,
loc: []float64{6, 7},
logProb: -3.503979321496947,
},
{
dist: dist2,
loc: []float64{1, 2},
logProb: -7.075407892925519,
},
})
}
func TestMarginal(t *testing.T) {
for _, test := range []struct {
mu []float64
sigma *mat64.SymDense
marginal []int
}{
{
mu: []float64{2, 3, 4},
sigma: mat64.NewSymDense(3, []float64{2, 0.5, 3, 0.5, 1, 0.6, 3, 0.6, 10}),
marginal: []int{0},
},
{
mu: []float64{2, 3, 4},
sigma: mat64.NewSymDense(3, []float64{2, 0.5, 3, 0.5, 1, 0.6, 3, 0.6, 10}),
marginal: []int{0, 2},
},
{
mu: []float64{2, 3, 4, 5},
sigma: mat64.NewSymDense(4, []float64{2, 0.5, 3, 0.1, 0.5, 1, 0.6, 0.2, 3, 0.6, 10, 0.3, 0.1, 0.2, 0.3, 3}),
marginal: []int{0, 3},
},
} {
normal, ok := NewNormal(test.mu, test.sigma, nil)
if !ok {
t.Fatalf("Bad test, covariance matrix not positive definite")
}
marginal, ok := normal.MarginalNormal(test.marginal, nil)
if !ok {
t.Fatalf("Bad test, marginal matrix not positive definite")
}
dim := normal.Dim()
nSamples := 1000000
samps := mat64.NewDense(nSamples, dim, nil)
for i := 0; i < nSamples; i++ {
normal.Rand(samps.RawRowView(i))
}
estMean := make([]float64, dim)
for i := range estMean {
estMean[i] = stat.Mean(mat64.Col(nil, i, samps), nil)
}
for i, v := range test.marginal {
if math.Abs(marginal.mu[i]-estMean[v]) > 1e-2 {
t.Errorf("Mean mismatch: want: %v, got %v", estMean[v], marginal.mu[i])
}
}
marginalCov := marginal.CovarianceMatrix(nil)
estCov := stat.CovarianceMatrix(nil, samps, nil)
for i, v1 := range test.marginal {
for j, v2 := range test.marginal {
c := marginalCov.At(i, j)
ec := estCov.At(v1, v2)
if math.Abs(c-ec) > 5e-2 {
t.Errorf("Cov mismatch element i = %d, j = %d: want: %v, got %v", i, j, c, ec)
}
}
}
}
}
func TestCovarianceMatrix(t *testing.T) {
for _, test := range []struct {
mu []float64
sigma *mat64.SymDense
}{
{
mu: []float64{2, 3, 4},
sigma: mat64.NewSymDense(3, []float64{1, 0.5, 3, 0.5, 8, -1, 3, -1, 15}),
},
} {
normal, ok := NewNormal(test.mu, test.sigma, nil)
if !ok {
t.Fatalf("Bad test, covariance matrix not positive definite")
}
cov := normal.CovarianceMatrix(nil)
if !mat64.EqualApprox(cov, test.sigma, 1e-14) {
t.Errorf("Covariance mismatch with nil input")
}
dim := test.sigma.Symmetric()
cov = mat64.NewSymDense(dim, nil)
normal.CovarianceMatrix(cov)
if !mat64.EqualApprox(cov, test.sigma, 1e-14) {
t.Errorf("Covariance mismatch with supplied input")
}
}
}
func (b *BFGS) InitDirection(loc *Location, dir []float64) (stepSize float64) {
dim := len(loc.X)
b.dim = dim
b.first = true
x := mat64.NewVector(dim, loc.X)
grad := mat64.NewVector(dim, loc.Gradient)
b.x.CloneVec(x)
b.grad.CloneVec(grad)
b.y.Reset()
b.s.Reset()
b.tmp.Reset()
if b.invHess == nil || cap(b.invHess.RawSymmetric().Data) < dim*dim {
b.invHess = mat64.NewSymDense(dim, nil)
} else {
b.invHess = mat64.NewSymDense(dim, b.invHess.RawSymmetric().Data[:dim*dim])
}
// The values of the inverse Hessian are initialized in the first call to
// NextDirection.
// Initial direction is just negative of the gradient because the Hessian
// is an identity matrix.
d := mat64.NewVector(dim, dir)
d.ScaleVec(-1, grad)
return 1 / mat64.Norm(d, 2)
}
func (b *BFGS) InitDirection(loc *Location, dir []float64) (stepSize float64) {
dim := len(loc.X)
b.dim = dim
b.x = resize(b.x, dim)
copy(b.x, loc.X)
b.grad = resize(b.grad, dim)
copy(b.grad, loc.Gradient)
b.y = resize(b.y, dim)
b.s = resize(b.s, dim)
b.tmp = resize(b.tmp, dim)
b.yVec = mat64.NewVector(dim, b.y)
b.sVec = mat64.NewVector(dim, b.s)
b.tmpVec = mat64.NewVector(dim, b.tmp)
if b.invHess == nil || cap(b.invHess.RawSymmetric().Data) < dim*dim {
b.invHess = mat64.NewSymDense(dim, nil)
} else {
b.invHess = mat64.NewSymDense(dim, b.invHess.RawSymmetric().Data[:dim*dim])
}
// The values of the hessian are initialized in the first call to NextDirection
// initial direcion is just negative of gradient because the hessian is 1
copy(dir, loc.Gradient)
floats.Scale(-1, dir)
b.first = true
return 1 / floats.Norm(dir, 2)
}
func newMargLikeMemory(hyper, outputs int) *margLikeMemory {
m := &margLikeMemory{
lastX: make([]float64, hyper),
k: mat64.NewSymDense(outputs, nil),
chol: &mat64.Cholesky{},
alpha: mat64.NewVector(outputs, nil),
tmp: mat64.NewVector(1, nil),
dKdTheta: make([]*mat64.SymDense, hyper),
kInvDK: mat64.NewDense(outputs, outputs, nil),
}
for i := 0; i < hyper; i++ {
m.dKdTheta[i] = mat64.NewSymDense(outputs, nil)
}
return m
}
// Cov returns the covariance between a set of data points based on the current
// GP fit.
func (g *GP) Cov(m *mat64.SymDense, x mat64.Matrix) *mat64.SymDense {
if m != nil {
// TODO(btracey): Make this k**
panic("resuing m not coded")
}
// The joint covariance matrix is
// K(x_*, k_*) - k(x_*, x) k(x,x)^-1 k(x, x*)
nSamp, nDim := x.Dims()
if nDim != g.inputDim {
panic(badInputLength)
}
// Compute K(x_*, x) K(x, x)^-1 K(x, x_*)
kstar := g.formKStar(x)
var tmp mat64.Dense
tmp.SolveCholesky(g.cholK, kstar)
var tmp2 mat64.Dense
tmp2.Mul(kstar.T(), &tmp)
// Compute k(x_*, x_*) and perform the subtraction.
kstarstar := mat64.NewSymDense(nSamp, nil)
for i := 0; i < nSamp; i++ {
for j := i; j < nSamp; j++ {
v := g.kernel.Distance(mat64.Row(nil, i, x), mat64.Row(nil, j, x))
if i == j {
v += g.noise
}
kstarstar.SetSym(i, j, v-tmp2.At(i, j))
}
}
return kstarstar
}
func TestMetropolisHastings(t *testing.T) {
// Test by finding the expected value of a normal distribution.
dim := 3
target, ok := randomNormal(dim)
if !ok {
t.Fatal("bad test, sigma not pos def")
}
sigmaImp := mat64.NewSymDense(dim, nil)
for i := 0; i < dim; i++ {
sigmaImp.SetSym(i, i, 0.25)
}
proposal, ok := NewProposalNormal(sigmaImp, nil)
if !ok {
t.Fatal("bad test, sigma not pos def")
}
nSamples := 1000000
burnin := 5000
batch := mat64.NewDense(nSamples, dim, nil)
initial := make([]float64, dim)
MetropolisHastings(batch, initial, target, proposal, nil)
batch = batch.View(burnin, 0, nSamples-burnin, dim).(*mat64.Dense)
compareNormal(t, target, batch, nil)
}
func TestImportance(t *testing.T) {
// Test by finding the expected value of a multi-variate normal.
dim := 3
target, ok := randomNormal(dim)
if !ok {
t.Fatal("bad test, sigma not pos def")
}
muImp := make([]float64, dim)
sigmaImp := mat64.NewSymDense(dim, nil)
for i := 0; i < dim; i++ {
sigmaImp.SetSym(i, i, 3)
}
proposal, ok := distmv.NewNormal(muImp, sigmaImp, nil)
if !ok {
t.Fatal("bad test, sigma not pos def")
}
nSamples := 100000
batch := mat64.NewDense(nSamples, dim, nil)
weights := make([]float64, nSamples)
Importance(batch, weights, target, proposal)
compareNormal(t, target, batch, weights)
}
// NewNormal creates a new Normal with the given mean and covariance matrix.
// NewNormal panics if len(mu) == 0, or if len(mu) != sigma.N. If the covariance
// matrix is not positive-definite, the returned boolean is false.
func NewNormal(mu []float64, sigma mat64.Symmetric, src *rand.Rand) (*Normal, bool) {
if len(mu) == 0 {
panic(badZeroDimension)
}
dim := sigma.Symmetric()
if dim != len(mu) {
panic(badSizeMismatch)
}
n := &Normal{
src: src,
dim: dim,
mu: make([]float64, dim),
sigma: mat64.NewSymDense(dim, nil),
chol: mat64.NewTriDense(dim, true, nil),
}
copy(n.mu, mu)
n.sigma.CopySym(sigma)
// TODO(btracey): Change this to the input Sigma, in case it is diagonal or
// banded.
ok := n.chol.Cholesky(n.sigma, true)
if !ok {
return nil, false
}
for i := 0; i < dim; i++ {
n.logSqrtDet += math.Log(n.chol.At(i, i))
}
return n, true
}
func benchmarkCovarianceMatrixInPlace(b *testing.B, m mat64.Matrix) {
_, c := m.Dims()
res := mat64.NewSymDense(c, nil)
b.ResetTimer()
for i := 0; i < b.N; i++ {
CovarianceMatrix(res, m, nil)
}
}
// getStartingLocation allocates and initializes the starting location for the minimization.
func getStartingLocation(p *Problem, method Method, initX []float64, stats *Stats, settings *Settings) (*Location, error) {
dim := len(initX)
loc := &Location{
X: make([]float64, dim),
}
copy(loc.X, initX)
if method.Needs().Gradient {
loc.Gradient = make([]float64, dim)
}
if method.Needs().Hessian {
loc.Hessian = mat64.NewSymDense(dim, nil)
}
if settings.UseInitialData {
loc.F = settings.InitialValue
if loc.Gradient != nil {
initG := settings.InitialGradient
if initG == nil {
panic("optimize: initial gradient is nil")
}
if len(initG) != dim {
panic("optimize: initial gradient size mismatch")
}
copy(loc.Gradient, initG)
}
if loc.Hessian != nil {
initH := settings.InitialHessian
if initH == nil {
panic("optimize: initial Hessian is nil")
}
if initH.Symmetric() != dim {
panic("optimize: initial Hessian size mismatch")
}
loc.Hessian.CopySym(initH)
}
} else {
eval := FuncEvaluation
if loc.Gradient != nil {
eval |= GradEvaluation
}
if loc.Hessian != nil {
eval |= HessEvaluation
}
x := make([]float64, len(loc.X))
evaluate(p, loc, eval, stats, x)
}
if math.IsInf(loc.F, 1) || math.IsNaN(loc.F) {
return loc, ErrFunc(loc.F)
}
for i, v := range loc.Gradient {
if math.IsInf(v, 0) || math.IsNaN(v) {
return loc, ErrGrad{Grad: v, Index: i}
}
}
return loc, nil
}
// NewUndirectedDenseGraph creates an undirected dense graph with n nodes.
// If passable is true all pairs of nodes will be connected by an edge
// with unit cost, otherwise every node will start unconnected with
// the cost specified by absent.
func NewUndirectedDenseGraph(n int, passable bool, absent float64) *UndirectedDenseGraph {
mat := make([]float64, n*n)
v := 1.
if !passable {
v = absent
}
for i := range mat {
mat[i] = v
}
return &UndirectedDenseGraph{mat: mat64.NewSymDense(n, mat), absent: absent}
}
// CovarianceMatrix returns the covariance matrix of the distribution. Upon
// return, the value at element {i, j} of the covariance matrix is equal to
// the covariance of the i^th and j^th variables.
// covariance(i, j) = E[(x_i - E[x_i])(x_j - E[x_j])]
// If the input matrix is nil a new matrix is allocated, otherwise the result
// is stored in-place into the input.
func (n *Normal) CovarianceMatrix(s *mat64.SymDense) *mat64.SymDense {
if s == nil {
s = mat64.NewSymDense(n.Dim(), nil)
}
sn := s.Symmetric()
if sn != n.Dim() {
panic("normal: input matrix size mismatch")
}
n.setSigma()
s.CopySym(n.sigma)
return s
}
func ExampleCholeskySymRankOne() {
a := mat64.NewSymDense(4, []float64{
1, 1, 1, 1,
0, 2, 3, 4,
0, 0, 6, 10,
0, 0, 0, 20,
})
fmt.Printf("A = %0.4v\n", mat64.Formatted(a, mat64.Prefix(" ")))
// Compute the Cholesky factorization.
var chol mat64.Cholesky
if ok := chol.Factorize(a); !ok {
fmt.Println("matrix a is not positive definite.")
}
x := mat64.NewVector(4, []float64{0, 0, 0, 1})
fmt.Printf("\nx = %0.4v\n", mat64.Formatted(x, mat64.Prefix(" ")))
// Rank-1 update the factorization.
chol.SymRankOne(&chol, 1, x)
// Rank-1 update the matrix a.
a.SymRankOne(a, 1, x)
var au mat64.SymDense
au.FromCholesky(&chol)
// Print the matrix that was updated directly.
fmt.Printf("\nA' = %0.4v\n", mat64.Formatted(a, mat64.Prefix(" ")))
// Print the matrix recovered from the factorization.
fmt.Printf("\nU'^T * U' = %0.4v\n", mat64.Formatted(&au, mat64.Prefix(" ")))
// Output:
// A = ⎡ 1 1 1 1⎤
// ⎢ 1 2 3 4⎥
// ⎢ 1 3 6 10⎥
// ⎣ 1 4 10 20⎦
//
// x = ⎡0⎤
// ⎢0⎥
// ⎢0⎥
// ⎣1⎦
//
// A' = ⎡ 1 1 1 1⎤
// ⎢ 1 2 3 4⎥
// ⎢ 1 3 6 10⎥
// ⎣ 1 4 10 21⎦
//
// U'^T * U' = ⎡ 1 1 1 1⎤
// ⎢ 1 2 3 4⎥
// ⎢ 1 3 6 10⎥
// ⎣ 1 4 10 21⎦
}
func TestNormRand(t *testing.T) {
for _, test := range []struct {
mean []float64
cov []float64
}{
{
mean: []float64{0, 0},
cov: []float64{
1, 0,
0, 1,
},
},
{
mean: []float64{0, 0},
cov: []float64{
1, 0.9,
0.9, 1,
},
},
{
mean: []float64{6, 7},
cov: []float64{
5, 0.9,
0.9, 2,
},
},
} {
dim := len(test.mean)
cov := mat64.NewSymDense(dim, test.cov)
n, ok := NewNormal(test.mean, cov, nil)
if !ok {
t.Errorf("bad covariance matrix")
}
nSamples := 1000000
samps := mat64.NewDense(nSamples, dim, nil)
for i := 0; i < nSamples; i++ {
n.Rand(samps.RawRowView(i))
}
estMean := make([]float64, dim)
for i := range estMean {
estMean[i] = stat.Mean(mat64.Col(nil, i, samps), nil)
}
if !floats.EqualApprox(estMean, test.mean, 1e-2) {
t.Errorf("Mean mismatch: want: %v, got %v", test.mean, estMean)
}
estCov := stat.CovarianceMatrix(nil, samps, nil)
if !mat64.EqualApprox(estCov, cov, 1e-2) {
t.Errorf("Cov mismatch: want: %v, got %v", cov, estCov)
}
}
}
func BenchmarkCovToCorr(b *testing.B) {
// generate a 10x10 covariance matrix
m := randMat(small, small)
c := CovarianceMatrix(nil, m, nil)
cc := mat64.NewSymDense(c.Symmetric(), nil)
b.ResetTimer()
for i := 0; i < b.N; i++ {
b.StopTimer()
cc.CopySym(c)
b.StartTimer()
covToCorr(cc)
}
}
// CovarianceMatrix calculates a covariance matrix (also known as a
// variance-covariance matrix) from a matrix of data, using a two-pass
// algorithm.
//
// The weights must have length equal to the number of rows in
// input data matrix x. If cov is nil, then a new matrix with appropriate size will
// be constructed. If cov is not nil, it should have the same number of columns as the
// input data matrix x, and it will be used as the destination for the covariance
// data. Weights must not be negative.
func CovarianceMatrix(cov *mat64.SymDense, x mat64.Matrix, weights []float64) *mat64.SymDense {
// This is the matrix version of the two-pass algorithm. It doesn't use the
// additional floating point error correction that the Covariance function uses
// to reduce the impact of rounding during centering.
r, c := x.Dims()
if cov == nil {
cov = mat64.NewSymDense(c, nil)
} else if n := cov.Symmetric(); n != c {
panic(matrix.ErrShape)
}
var xt mat64.Dense
xt.Clone(x.T())
// Subtract the mean of each of the columns.
for i := 0; i < c; i++ {
v := xt.RawRowView(i)
// This will panic with ErrShape if len(weights) != len(v), so
// we don't have to check the size later.
mean := Mean(v, weights)
floats.AddConst(-mean, v)
}
if weights == nil {
// Calculate the normalization factor
// scaled by the sample size.
cov.SymOuterK(1/(float64(r)-1), &xt)
return cov
}
// Multiply by the sqrt of the weights, so that multiplication is symmetric.
sqrtwts := make([]float64, r)
for i, w := range weights {
if w < 0 {
panic("stat: negative covariance matrix weights")
}
sqrtwts[i] = math.Sqrt(w)
}
// Weight the rows.
for i := 0; i < c; i++ {
v := xt.RawRowView(i)
floats.Mul(v, sqrtwts)
}
// Calculate the normalization factor
// scaled by the weighted sample size.
cov.SymOuterK(1/(floats.Sum(weights)-1), &xt)
return cov
}
func copyLocation(dst, src *Location) {
dst.X = resize(dst.X, len(src.X))
copy(dst.X, src.X)
dst.F = src.F
dst.Gradient = resize(dst.Gradient, len(src.Gradient))
copy(dst.Gradient, src.Gradient)
if src.Hessian != nil {
if dst.Hessian == nil || dst.Hessian.Symmetric() != len(src.X) {
dst.Hessian = mat64.NewSymDense(len(src.X), nil)
}
dst.Hessian.CopySym(src.Hessian)
}
}
// NewUndirectedMatrix creates an undirected dense graph with n nodes.
// All edges are initialized with the weight given by init. The self parameter
// specifies the cost of self connection, and absent specifies the weight
// returned for absent edges.
func NewUndirectedMatrix(n int, init, self, absent float64) *UndirectedMatrix {
mat := make([]float64, n*n)
if init != 0 {
for i := range mat {
mat[i] = init
}
}
for i := 0; i < len(mat); i += n + 1 {
mat[i] = self
}
return &UndirectedMatrix{
mat: mat64.NewSymDense(n, mat),
self: self,
absent: absent,
}
}
func BenchmarkCorrToCov(b *testing.B) {
// generate a 10x10 correlation matrix
m := randMat(small, small)
c := CorrelationMatrix(nil, m, nil)
cc := mat64.NewSymDense(c.Symmetric(), nil)
sigma := make([]float64, small)
for i := range sigma {
sigma[i] = 2
}
b.ResetTimer()
for i := 0; i < b.N; i++ {
b.StopTimer()
cc.CopySym(c)
b.StartTimer()
corrToCov(cc, sigma)
}
}
func ExampleSymDense_SubsetSym() {
n := 5
s := mat64.NewSymDense(5, nil)
count := 1.0
for i := 0; i < n; i++ {
for j := i; j < n; j++ {
s.SetSym(i, j, count)
count++
}
}
fmt.Println("Original matrix:")
fmt.Printf("%0.4v\n\n", mat64.Formatted(s))
// Take the subset {0, 2, 4}
var sub mat64.SymDense
sub.SubsetSym(s, []int{0, 2, 4})
fmt.Println("Subset {0, 2, 4}")
fmt.Printf("%0.4v\n\n", mat64.Formatted(&sub))
// Take the subset {0, 0, 4}
sub.SubsetSym(s, []int{0, 0, 4})
fmt.Println("Subset {0, 0, 4}")
fmt.Printf("%0.4v\n\n", mat64.Formatted(&sub))
// Output:
// Original matrix:
// ⎡ 1 2 3 4 5⎤
// ⎢ 2 6 7 8 9⎥
// ⎢ 3 7 10 11 12⎥
// ⎢ 4 8 11 13 14⎥
// ⎣ 5 9 12 14 15⎦
//
// Subset {0, 2, 4}
// ⎡ 1 3 5⎤
// ⎢ 3 10 12⎥
// ⎣ 5 12 15⎦
//
// Subset {0, 0, 4}
// ⎡ 1 1 5⎤
// ⎢ 1 1 5⎥
// ⎣ 5 5 15⎦
}
func TestRejection(t *testing.T) {
// Test by finding the expected value of a uniform.
dim := 3
bounds := make([]distmv.Bound, dim)
for i := 0; i < dim; i++ {
min := rand.NormFloat64()
max := rand.NormFloat64()
if min > max {
min, max = max, min
}
bounds[i].Min = min
bounds[i].Max = max
}
target := distmv.NewUniform(bounds, nil)
mu := target.Mean(nil)
muImp := make([]float64, dim)
sigmaImp := mat64.NewSymDense(dim, nil)
for i := 0; i < dim; i++ {
sigmaImp.SetSym(i, i, 6)
}
proposal, ok := distmv.NewNormal(muImp, sigmaImp, nil)
if !ok {
t.Fatal("bad test, sigma not pos def")
}
nSamples := 1000
batch := mat64.NewDense(nSamples, dim, nil)
weights := make([]float64, nSamples)
_, ok = Rejection(batch, target, proposal, 1000, nil)
if !ok {
t.Error("Bad test, nan samples")
}
for i := 0; i < dim; i++ {
col := mat64.Col(nil, i, batch)
ev := stat.Mean(col, weights)
if math.Abs(ev-mu[i]) > 1e-2 {
t.Errorf("Mean mismatch: Want %v, got %v", mu[i], ev)
}
}
}
// New creates a new GP with the given input dimension, the given
// kernel function, and output noise parameter. Output dim must be one.
func New(inputDim int, kernel Kernel, noise float64) *GP {
if inputDim <= 0 {
panic("gp: non-positive inputDim")
}
if kernel == nil {
panic("gp: nil kernel")
}
if !(noise >= 0) {
panic("gp: negative noise") // also handles NaN.
}
return &GP{
kernel: kernel,
noise: noise,
inputDim: inputDim,
mean: 0,
std: 1,
inputs: &mat64.Dense{},
outputs: make([]float64, 0),
k: mat64.NewSymDense(0, nil),
sigInvY: &mat64.Vector{},
cholK: &mat64.Cholesky{},
}
}
func TestConditionNormal(t *testing.T) {
// Uncorrelated values shouldn't influence the updated values.
for _, test := range []struct {
mu []float64
sigma *mat64.SymDense
observed []int
values []float64
newMu []float64
newSigma *mat64.SymDense
}{
{
mu: []float64{2, 3},
sigma: mat64.NewSymDense(2, []float64{2, 0, 0, 5}),
observed: []int{0},
values: []float64{10},
newMu: []float64{3},
newSigma: mat64.NewSymDense(1, []float64{5}),
},
{
mu: []float64{2, 3},
sigma: mat64.NewSymDense(2, []float64{2, 0, 0, 5}),
observed: []int{1},
values: []float64{10},
newMu: []float64{2},
newSigma: mat64.NewSymDense(1, []float64{2}),
},
{
mu: []float64{2, 3, 4},
sigma: mat64.NewSymDense(3, []float64{2, 0, 0, 0, 5, 0, 0, 0, 10}),
observed: []int{1},
values: []float64{10},
newMu: []float64{2, 4},
newSigma: mat64.NewSymDense(2, []float64{2, 0, 0, 10}),
},
{
mu: []float64{2, 3, 4},
sigma: mat64.NewSymDense(3, []float64{2, 0, 0, 0, 5, 0, 0, 0, 10}),
observed: []int{0, 1},
values: []float64{10, 15},
newMu: []float64{4},
newSigma: mat64.NewSymDense(1, []float64{10}),
},
{
mu: []float64{2, 3, 4, 5},
sigma: mat64.NewSymDense(4, []float64{2, 0.5, 0, 0, 0.5, 5, 0, 0, 0, 0, 10, 2, 0, 0, 2, 3}),
observed: []int{0, 1},
values: []float64{10, 15},
newMu: []float64{4, 5},
newSigma: mat64.NewSymDense(2, []float64{10, 2, 2, 3}),
},
} {
normal, ok := NewNormal(test.mu, test.sigma, nil)
if !ok {
t.Fatalf("Bad test, original sigma not positive definite")
}
newNormal, ok := normal.ConditionNormal(test.observed, test.values, nil)
if !ok {
t.Fatalf("Bad test, update failure")
}
if !floats.EqualApprox(test.newMu, newNormal.mu, 1e-12) {
t.Errorf("Updated mean mismatch. Want %v, got %v.", test.newMu, newNormal.mu)
}
var sigma mat64.SymDense
sigma.FromCholesky(&newNormal.chol)
if !mat64.EqualApprox(test.newSigma, &sigma, 1e-12) {
t.Errorf("Updated sigma mismatch\n.Want:\n% v\nGot:\n% v\n", test.newSigma, sigma)
}
}
// Test bivariate case where the update rule is analytic
for _, test := range []struct {
mu []float64
std []float64
rho float64
value float64
}{
{
mu: []float64{2, 3},
std: []float64{3, 5},
rho: 0.9,
value: 1000,
},
{
mu: []float64{2, 3},
std: []float64{3, 5},
rho: -0.9,
value: 1000,
},
} {
std := test.std
rho := test.rho
sigma := mat64.NewSymDense(2, []float64{std[0] * std[0], std[0] * std[1] * rho, std[0] * std[1] * rho, std[1] * std[1]})
normal, ok := NewNormal(test.mu, sigma, nil)
if !ok {
t.Fatalf("Bad test, original sigma not positive definite")
}
newNormal, ok := normal.ConditionNormal([]int{1}, []float64{test.value}, nil)
if !ok {
t.Fatalf("Bad test, update failed")
}
var newSigma mat64.SymDense
newSigma.FromCholesky(&newNormal.chol)
trueMean := test.mu[0] + rho*(std[0]/std[1])*(test.value-test.mu[1])
if math.Abs(trueMean-newNormal.mu[0]) > 1e-14 {
t.Errorf("Mean mismatch. Want %v, got %v", trueMean, newNormal.mu[0])
}
trueVar := (1 - rho*rho) * std[0] * std[0]
if math.Abs(trueVar-newSigma.At(0, 0)) > 1e-14 {
t.Errorf("Std mismatch. Want %v, got %v", trueMean, newNormal.mu[0])
}
}
// Test via sampling.
for _, test := range []struct {
mu []float64
sigma *mat64.SymDense
observed []int
unobserved []int
value []float64
}{
// The indices in unobserved must be in ascending order for this test.
{
mu: []float64{2, 3, 4},
sigma: mat64.NewSymDense(3, []float64{2, 0.5, 3, 0.5, 1, 0.6, 3, 0.6, 10}),
observed: []int{0},
unobserved: []int{1, 2},
value: []float64{1.9},
},
{
mu: []float64{2, 3, 4, 5},
sigma: mat64.NewSymDense(4, []float64{2, 0.5, 3, 0.1, 0.5, 1, 0.6, 0.2, 3, 0.6, 10, 0.3, 0.1, 0.2, 0.3, 3}),
observed: []int{0, 3},
unobserved: []int{1, 2},
value: []float64{1.9, 2.9},
},
} {
totalSamp := 4000000
var nSamp int
samples := mat64.NewDense(totalSamp, len(test.mu), nil)
normal, ok := NewNormal(test.mu, test.sigma, nil)
if !ok {
t.Errorf("bad test")
}
sample := make([]float64, len(test.mu))
for i := 0; i < totalSamp; i++ {
normal.Rand(sample)
isClose := true
for i, v := range test.observed {
if math.Abs(sample[v]-test.value[i]) > 1e-1 {
isClose = false
break
}
}
if isClose {
samples.SetRow(nSamp, sample)
nSamp++
}
}
if nSamp < 100 {
t.Errorf("bad test, not enough samples")
continue
}
samples = samples.View(0, 0, nSamp, len(test.mu)).(*mat64.Dense)
// Compute mean and covariance matrix.
estMean := make([]float64, len(test.mu))
for i := range estMean {
estMean[i] = stat.Mean(mat64.Col(nil, i, samples), nil)
}
estCov := stat.CovarianceMatrix(nil, samples, nil)
// Compute update rule.
newNormal, ok := normal.ConditionNormal(test.observed, test.value, nil)
if !ok {
t.Fatalf("Bad test, update failure")
}
var subEstMean []float64
for _, v := range test.unobserved {
subEstMean = append(subEstMean, estMean[v])
}
subEstCov := mat64.NewSymDense(len(test.unobserved), nil)
for i := 0; i < len(test.unobserved); i++ {
for j := i; j < len(test.unobserved); j++ {
subEstCov.SetSym(i, j, estCov.At(test.unobserved[i], test.unobserved[j]))
}
}
for i, v := range subEstMean {
if math.Abs(newNormal.mu[i]-v) > 5e-2 {
t.Errorf("Mean mismatch. Want %v, got %v.", newNormal.mu[i], v)
}
}
var sigma mat64.SymDense
sigma.FromCholesky(&newNormal.chol)
if !mat64.EqualApprox(&sigma, subEstCov, 1e-1) {
t.Errorf("Covariance mismatch. Want:\n%0.8v\nGot:\n%0.8v\n", subEstCov, sigma)
}
}
}
func resizeSymDense(m *mat64.SymDense, dim int) *mat64.SymDense {
if m == nil || cap(m.RawSymmetric().Data) < dim*dim {
return mat64.NewSymDense(dim, nil)
}
return mat64.NewSymDense(dim, m.RawSymmetric().Data[:dim*dim])
}
func TestMetropolisHastingser(t *testing.T) {
for seed, test := range []struct {
dim, burnin, rate, samples int
}{
{3, 10, 1, 1},
{3, 10, 2, 1},
{3, 10, 1, 2},
{3, 10, 3, 2},
{3, 10, 7, 4},
{3, 10, 7, 4},
{3, 11, 51, 103},
{3, 11, 103, 51},
{3, 51, 11, 103},
{3, 51, 103, 11},
{3, 103, 11, 51},
{3, 103, 51, 11},
} {
dim := test.dim
initial := make([]float64, dim)
target, ok := randomNormal(dim)
if !ok {
t.Fatal("bad test, sigma not pos def")
}
sigmaImp := mat64.NewSymDense(dim, nil)
for i := 0; i < dim; i++ {
sigmaImp.SetSym(i, i, 0.25)
}
proposal, ok := NewProposalNormal(sigmaImp, nil)
if !ok {
t.Fatal("bad test, sigma not pos def")
}
// Test the Metropolis Hastingser by generating all the samples, then generating
// the same samples with a burnin and rate.
rand.Seed(int64(seed))
mh := MetropolisHastingser{
Initial: initial,
Target: target,
Proposal: proposal,
Src: nil,
BurnIn: 0,
Rate: 0,
}
samples := test.samples
burnin := test.burnin
rate := test.rate
fullBatch := mat64.NewDense(1+burnin+rate*(samples-1), dim, nil)
mh.Sample(fullBatch)
mh = MetropolisHastingser{
Initial: initial,
Target: target,
Proposal: proposal,
Src: nil,
BurnIn: burnin,
Rate: rate,
}
rand.Seed(int64(seed))
batch := mat64.NewDense(samples, dim, nil)
mh.Sample(batch)
same := true
count := burnin
for i := 0; i < samples; i++ {
if !floats.Equal(batch.RawRowView(i), fullBatch.RawRowView(count)) {
fmt.Println("sample ", i, "is different")
same = false
break
}
count += rate
}
if !same {
fmt.Printf("%v\n", mat64.Formatted(batch))
fmt.Printf("%v\n", mat64.Formatted(fullBatch))
t.Errorf("sampling mismatch: dim = %v, burnin = %v, rate = %v, samples = %v", dim, burnin, rate, samples)
}
}
}
// AddBatch adds a set training points to the Gp. This call updates internal
// values needed for prediction, so it is more efficient to add samples
// as a batch.
func (g *GP) AddBatch(x mat64.Matrix, y []float64) error {
// Note: The outputs are stored scaled to have a mean of zero and a variance
// of 1.
// Verify input parameters
rx, cx := x.Dims()
ry := len(y)
if rx != ry {
panic(badInOut)
}
if cx != g.inputDim {
panic(badInputLength)
}
nSamples := len(g.outputs)
// Append the new data to the list of stored data.
inputs := mat64.NewDense(rx+nSamples, g.inputDim, nil)
inputs.Copy(g.inputs)
inputs.View(nSamples, 0, rx, g.inputDim).(*mat64.Dense).Copy(x)
g.inputs = inputs
// Rescale the output data to its original value, append the new data, and
// then rescale to have mean 0 and variance of 1.
for i, v := range g.outputs {
g.outputs[i] = v*g.std + g.mean
}
g.outputs = append(g.outputs, y...)
g.mean = stat.Mean(g.outputs, nil)
g.std = stat.StdDev(g.outputs, nil)
for i, v := range g.outputs {
g.outputs[i] = (v - g.mean) / g.std
}
// Add to the kernel matrix.
k := mat64.NewSymDense(rx+nSamples, nil)
k.CopySym(g.k)
g.k = k
// Compute the kernel with the new points and the old points
for i := 0; i < nSamples; i++ {
for j := nSamples; j < rx+nSamples; j++ {
v := g.kernel.Distance(g.inputs.RawRowView(i), g.inputs.RawRowView(j))
g.k.SetSym(i, j, v)
}
}
// Compute the kernel with the new points and themselves
for i := nSamples; i < rx+nSamples; i++ {
for j := i; j < nSamples+rx; j++ {
v := g.kernel.Distance(g.inputs.RawRowView(i), g.inputs.RawRowView(j))
if i == j {
v += g.noise
}
g.k.SetSym(i, j, v)
}
}
// Cache necessary matrix results for computing predictions.
var chol mat64.Cholesky
ok := chol.Factorize(g.k)
if !ok {
return ErrSingular
}
g.cholK = &chol
g.sigInvY.Reset()
v := mat64.NewVector(len(g.outputs), g.outputs)
g.sigInvY.SolveCholeskyVec(g.cholK, v)
return nil
}
// setSigma computes and stores the covariance matrix of the distribution.
func (n *Normal) setSigma() {
n.once.Do(func() {
n.sigma = mat64.NewSymDense(n.Dim(), nil)
n.sigma.FromCholesky(&n.chol)
})
}
func TestMarginalSingle(t *testing.T) {
for _, test := range []struct {
mu []float64
sigma *mat64.SymDense
}{
{
mu: []float64{2, 3, 4},
sigma: mat64.NewSymDense(3, []float64{2, 0.5, 3, 0.5, 1, 0.6, 3, 0.6, 10}),
},
{
mu: []float64{2, 3, 4, 5},
sigma: mat64.NewSymDense(4, []float64{2, 0.5, 3, 0.1, 0.5, 1, 0.6, 0.2, 3, 0.6, 10, 0.3, 0.1, 0.2, 0.3, 3}),
},
} {
normal, ok := NewNormal(test.mu, test.sigma, nil)
if !ok {
t.Fatalf("Bad test, covariance matrix not positive definite")
}
// Verify with nil Sigma.
normal.sigma = nil
for i, mean := range test.mu {
norm := normal.MarginalNormalSingle(i, nil)
if norm.Mean() != mean {
t.Errorf("Mean mismatch nil Sigma, idx %v: want %v, got %v.", i, mean, norm.Mean())
}
std := math.Sqrt(test.sigma.At(i, i))
if math.Abs(norm.StdDev()-std) > 1e-14 {
t.Errorf("StdDev mismatch nil Sigma, idx %v: want %v, got %v.", i, std, norm.StdDev())
}
}
// Verify with non-nil Sigma.
normal.setSigma()
for i, mean := range test.mu {
norm := normal.MarginalNormalSingle(i, nil)
if norm.Mean() != mean {
t.Errorf("Mean mismatch non-nil Sigma, idx %v: want %v, got %v.", i, mean, norm.Mean())
}
std := math.Sqrt(test.sigma.At(i, i))
if math.Abs(norm.StdDev()-std) > 1e-14 {
t.Errorf("StdDev mismatch non-nil Sigma, idx %v: want %v, got %v.", i, std, norm.StdDev())
}
}
}
// Test matching with TestMarginal.
rnd := rand.New(rand.NewSource(1))
for cas := 0; cas < 10; cas++ {
dim := rnd.Intn(10) + 1
mu := make([]float64, dim)
for i := range mu {
mu[i] = rnd.Float64()
}
x := make([]float64, dim*dim)
for i := range x {
x[i] = rnd.Float64()
}
mat := mat64.NewDense(dim, dim, x)
var sigma mat64.SymDense
sigma.SymOuterK(1, mat)
normal, ok := NewNormal(mu, &sigma, nil)
if !ok {
t.Fatal("bad test")
}
for i := 0; i < dim; i++ {
single := normal.MarginalNormalSingle(i, nil)
mult, ok := normal.MarginalNormal([]int{i}, nil)
if !ok {
t.Fatal("bad test")
}
if math.Abs(single.Mean()-mult.Mean(nil)[0]) > 1e-14 {
t.Errorf("Mean mismatch")
}
if math.Abs(single.Variance()-mult.CovarianceMatrix(nil).At(0, 0)) > 1e-14 {
t.Errorf("Variance mismatch")
}
}
}
}
