mat: rename Symmetric method to SymmetricDim

diff/fd/hessian.go

@@ -25,7 +25,7 @@ func Hessian(dst *mat.SymDense, f func(x []float64) float64, x []float64, settin
 	n := len(x)
 	if dst.IsEmpty() {
 		*dst = *(dst.GrowSym(n).(*mat.SymDense))
-	} else if dst.Symmetric() != n {
+	} else if dst.SymmetricDim() != n {
 		panic("hessian: dst size mismatch")
 	}
 	dst.Zero()
@@ -111,7 +111,7 @@ func hessianSerial(dst *mat.SymDense, f func(x []float64) float64, x []float64,
 }
 
 func hessianConcurrent(dst *mat.SymDense, nWorkers, evals int, f func(x []float64) float64, x []float64, stencil []Point, step float64, originKnown bool, originValue float64) {
-	n := dst.Symmetric()
+	n := dst.SymmetricDim()
 	type run struct {
 		i, j       int
 		iIdx, jIdx int

diff/fd/watson_test.go

@@ -86,7 +86,7 @@ func (Watson) Grad(grad, x []float64) {
 
 func (Watson) Hess(hess mat.MutableSymmetric, x []float64) {
 	dim := len(x)
-	if dim != hess.Symmetric() {
+	if dim != hess.SymmetricDim() {
 		panic("incorrect size of the Hessian")
 	}
 

graph/simple/dense_undirected_matrix.go

@@ -107,7 +107,7 @@ func (g *UndirectedMatrix) From(id int64) graph.Nodes {
 		return graph.Empty
 	}
 	var nodes []graph.Node
-	r := g.mat.Symmetric()
+	r := g.mat.SymmetricDim()
 	for i := 0; i < r; i++ {
 		if int64(i) == id {
 			continue
@@ -161,7 +161,7 @@ func (g *UndirectedMatrix) Nodes() graph.Nodes {
 		copy(nodes, g.nodes)
 		return iterator.NewOrderedNodes(nodes)
 	}
-	r := g.mat.Symmetric()
+	r := g.mat.SymmetricDim()
 	// Matrix graphs must have at least one node.
 	return iterator.NewImplicitNodes(0, r, newSimpleNode)
 }
@@ -263,6 +263,6 @@ func (g *UndirectedMatrix) WeightedEdges() graph.WeightedEdges {
 }
 
 func (g *UndirectedMatrix) has(id int64) bool {
-	r := g.mat.Symmetric()
+	r := g.mat.SymmetricDim()
 	return 0 <= id && id < int64(r)
 }

mat/basictypes_test.go

@@ -56,7 +56,7 @@ var _ Symmetric = &basicSymmetric{}
 func (m *basicSymmetric) At(r, c int) float64 { return (*SymDense)(m).At(r, c) }
 func (m *basicSymmetric) Dims() (r, c int)    { return (*SymDense)(m).Dims() }
 func (m *basicSymmetric) T() Matrix           { return m }
-func (m *basicSymmetric) Symmetric() int      { return (*SymDense)(m).Symmetric() }
+func (m *basicSymmetric) SymmetricDim() int   { return (*SymDense)(m).SymmetricDim() }
 
 type basicTriangular TriDense
 
@@ -87,7 +87,7 @@ func (m *basicSymBanded) Dims() (r, c int)        { return (*SymBandDense)(m).Di
 func (m *basicSymBanded) T() Matrix               { return m }
 func (m *basicSymBanded) Bandwidth() (kl, ku int) { return (*SymBandDense)(m).Bandwidth() }
 func (m *basicSymBanded) TBand() Banded           { return m }
-func (m *basicSymBanded) Symmetric() int          { return (*SymBandDense)(m).Symmetric() }
+func (m *basicSymBanded) SymmetricDim() int       { return (*SymBandDense)(m).SymmetricDim() }
 func (m *basicSymBanded) SymBand() (n, k int)     { return (*SymBandDense)(m).SymBand() }
 
 type basicTriBanded TriBandDense
@@ -112,7 +112,7 @@ func (m *basicDiagonal) At(r, c int) float64               { return (*DiagDense)
 func (m *basicDiagonal) Dims() (r, c int)                  { return (*DiagDense)(m).Dims() }
 func (m *basicDiagonal) T() Matrix                         { return Transpose{m} }
 func (m *basicDiagonal) Diag() int                         { return (*DiagDense)(m).Diag() }
-func (m *basicDiagonal) Symmetric() int                    { return (*DiagDense)(m).Symmetric() }
+func (m *basicDiagonal) SymmetricDim() int                 { return (*DiagDense)(m).SymmetricDim() }
 func (m *basicDiagonal) SymBand() (n, k int)               { return (*DiagDense)(m).SymBand() }
 func (m *basicDiagonal) Bandwidth() (kl, ku int)           { return (*DiagDense)(m).Bandwidth() }
 func (m *basicDiagonal) TBand() Banded                     { return TransposeBand{m} }

mat/cholesky.go

@@ -90,7 +90,7 @@ func (c *Cholesky) At(i, j int) float64 {
 	if !c.valid() {
 		panic(badCholesky)
 	}
-	n := c.Symmetric()
+	n := c.SymmetricDim()
 	if uint(i) >= uint(n) {
 		panic(ErrRowAccess)
 	}
@@ -110,9 +110,9 @@ func (c *Cholesky) T() Matrix {
 	return c
 }
 
-// Symmetric implements the Symmetric interface and returns the number of rows
+// SymmetricDim implements the Symmetric interface and returns the number of rows
 // in the matrix (this is also the number of columns).
-func (c *Cholesky) Symmetric() int {
+func (c *Cholesky) SymmetricDim() int {
 	r, _ := c.chol.Dims()
 	return r
 }
@@ -129,7 +129,7 @@ func (c *Cholesky) Cond() float64 {
 // whether the matrix is positive definite. If Factorize returns false, the
 // factorization must not be used.
 func (c *Cholesky) Factorize(a Symmetric) (ok bool) {
-	n := a.Symmetric()
+	n := a.SymmetricDim()
 	if c.chol == nil {
 		c.chol = NewTriDense(n, Upper, nil)
 	} else {
@@ -193,7 +193,7 @@ func (c *Cholesky) Clone(chol *Cholesky) {
 	if !chol.valid() {
 		panic(badCholesky)
 	}
-	n := chol.Symmetric()
+	n := chol.SymmetricDim()
 	if c.chol == nil {
 		c.chol = NewTriDense(n, Upper, nil)
 	} else {
@@ -377,7 +377,7 @@ func (c *Cholesky) ToSym(dst *SymDense) {
 	if dst.IsEmpty() {
 		dst.ReuseAsSym(n)
 	} else {
-		n2 := dst.Symmetric()
+		n2 := dst.SymmetricDim()
 		if n != n2 {
 			panic(ErrShape)
 		}
@@ -458,7 +458,7 @@ func (c *Cholesky) Scale(f float64, orig *Cholesky) {
 	if f <= 0 {
 		panic("cholesky: scaling by a non-positive constant")
 	}
-	n := orig.Symmetric()
+	n := orig.SymmetricDim()
 	if c.chol == nil {
 		c.chol = NewTriDense(n, Upper, nil)
 	} else if c.chol.mat.N != n {
@@ -480,7 +480,7 @@ func (c *Cholesky) Scale(f float64, orig *Cholesky) {
 // ExtendVecSym will panic if v.Len() != a.Symmetric()+1 or if a does not contain
 // a valid decomposition.
 func (c *Cholesky) ExtendVecSym(a *Cholesky, v Vector) (ok bool) {
-	n := a.Symmetric()
+	n := a.SymmetricDim()
 
 	if v.Len() != n+1 {
 		panic(badSliceLength)
@@ -549,7 +549,7 @@ func (c *Cholesky) SymRankOne(orig *Cholesky, alpha float64, x Vector) (ok bool)
 	if !orig.valid() {
 		panic(badCholesky)
 	}
-	n := orig.Symmetric()
+	n := orig.SymmetricDim()
 	if r, c := x.Dims(); r != n || c != 1 {
 		panic(ErrShape)
 	}
@@ -872,9 +872,9 @@ func (ch *BandCholesky) TBand() Banded {
 	return ch
 }
 
-// Symmetric implements the Symmetric interface and returns the number of rows
+// SymmetricDim implements the Symmetric interface and returns the number of rows
 // in the matrix (this is also the number of columns).
-func (ch *BandCholesky) Symmetric() int {
+func (ch *BandCholesky) SymmetricDim() int {
 	n, _ := ch.chol.Triangle()
 	return n
 }

mat/cholesky_test.go

@@ -89,8 +89,8 @@ func TestCholeskyAt(t *testing.T) {
 		if !ok {
 			t.Fatalf("Matrix not positive definite")
 		}
-		n := test.Symmetric()
-		cn := chol.Symmetric()
+		n := test.SymmetricDim()
+		cn := chol.SymmetricDim()
 		if cn != n {
 			t.Errorf("Cholesky size does not match. Got %d, want %d", cn, n)
 		}
@@ -516,7 +516,7 @@ func TestCholeskyExtendVecSym(t *testing.T) {
 			}),
 		},
 	} {
-		n := test.a.Symmetric()
+		n := test.a.SymmetricDim()
 		as := test.a.sliceSym(0, n-1)
 
 		// Compute the full factorization to use later (do the full factorization

mat/diagonal.go

@@ -54,10 +54,10 @@ type Diagonal interface {
 	Triangle() (int, TriKind)
 	TTri() Triangular
 
-	// Symmetric and SymBand are included in the Diagonal interface
+	// SymmetricDim and SymBand are included in the Diagonal interface
 	// to allow the use of Diagonal types in symmetric and banded symmetric
 	// functions respectively.
-	Symmetric() int
+	SymmetricDim() int
 	SymBand() (n, k int)
 
 	// TriBand and TTriBand are included in the Diagonal interface
@@ -137,8 +137,8 @@ func (d *DiagDense) Bandwidth() (kl, ku int) {
 	return 0, 0
 }
 
-// Symmetric implements the Symmetric interface.
-func (d *DiagDense) Symmetric() int {
+// SymmetricDim implements the Symmetric interface.
+func (d *DiagDense) SymmetricDim() int {
 	return d.mat.N
 }
 

mat/eigen.go

@@ -38,7 +38,7 @@ func (e *EigenSym) Factorize(a Symmetric, vectors bool) (ok bool) {
 	e.vectorsComputed = false
 	e.values = e.values[:]
 
-	n := a.Symmetric()
+	n := a.SymmetricDim()
 	sd := NewSymDense(n, nil)
 	sd.CopySym(a)
 

mat/list_test.go

@@ -765,7 +765,7 @@ func makeCopyOf(a Matrix) Matrix {
 		m.CloneFrom(a)
 		return returnAs(&m, t)
 	case *SymDense, *basicSymmetric:
-		n := t.(Symmetric).Symmetric()
+		n := t.(Symmetric).SymmetricDim()
 		m := NewSymDense(n, nil)
 		m.CopySym(t.(Symmetric))
 		return returnAs(m, t)

mat/symband.go

@@ -36,8 +36,8 @@ type SymBandDense struct {
 type SymBanded interface {
 	Banded
 
-	// Symmetric returns the number of rows/columns in the matrix.
-	Symmetric() int
+	// SymmetricDim returns the number of rows/columns in the matrix.
+	SymmetricDim() int
 
 	// SymBand returns the number of rows/columns in the matrix, and the size of
 	// the bandwidth.
@@ -115,8 +115,8 @@ func (s *SymBandDense) Dims() (r, c int) {
 	return s.mat.N, s.mat.N
 }
 
-// Symmetric returns the size of the receiver.
-func (s *SymBandDense) Symmetric() int {
+// SymmetricDim returns the size of the receiver.
+func (s *SymBandDense) SymmetricDim() int {
 	return s.mat.N
 }
 

mat/symmetric.go

@@ -37,8 +37,8 @@ type SymDense struct {
 // the element at {j, i}). Symmetric matrices are always square.
 type Symmetric interface {
 	Matrix
-	// Symmetric returns the number of rows/columns in the matrix.
-	Symmetric() int
+	// SymmetricDim returns the number of rows/columns in the matrix.
+	SymmetricDim() int
 }
 
 // A RawSymmetricer can return a view of itself as a BLAS Symmetric matrix.
@@ -100,9 +100,9 @@ func (s *SymDense) T() Matrix {
 	return s
 }
 
-// Symmetric implements the Symmetric interface and returns the number of rows
+// SymmetricDim implements the Symmetric interface and returns the number of rows
 // and columns in the matrix.
-func (s *SymDense) Symmetric() int {
+func (s *SymDense) SymmetricDim() int {
 	return s.mat.N
 }
 
@@ -234,7 +234,7 @@ func (s *SymDense) reuseAsZeroed(n int) {
 }
 
 func (s *SymDense) isolatedWorkspace(a Symmetric) (w *SymDense, restore func()) {
-	n := a.Symmetric()
+	n := a.SymmetricDim()
 	if n == 0 {
 		panic(ErrZeroLength)
 	}
@@ -258,8 +258,8 @@ func (s *SymDense) DiagView() Diagonal {
 }
 
 func (s *SymDense) AddSym(a, b Symmetric) {
-	n := a.Symmetric()
-	if n != b.Symmetric() {
+	n := a.SymmetricDim()
+	if n != b.SymmetricDim() {
 		panic(ErrShape)
 	}
 	s.reuseAsNonZeroed(n)
@@ -295,7 +295,7 @@ func (s *SymDense) AddSym(a, b Symmetric) {
 }
 
 func (s *SymDense) CopySym(a Symmetric) int {
-	n := a.Symmetric()
+	n := a.SymmetricDim()
 	n = min(n, s.mat.N)
 	if n == 0 {
 		return 0
@@ -325,7 +325,7 @@ func (s *SymDense) CopySym(a Symmetric) int {
 //  s = a + alpha * x * xᵀ
 func (s *SymDense) SymRankOne(a Symmetric, alpha float64, x Vector) {
 	n := x.Len()
-	if a.Symmetric() != n {
+	if a.SymmetricDim() != n {
 		panic(ErrShape)
 	}
 	s.reuseAsNonZeroed(n)
@@ -357,7 +357,7 @@ func (s *SymDense) SymRankOne(a Symmetric, alpha float64, x Vector) {
 // result into the receiver. If a is zero, see SymOuterK.
 //  s = a + alpha * x * x'
 func (s *SymDense) SymRankK(a Symmetric, alpha float64, x Matrix) {
-	n := a.Symmetric()
+	n := a.SymmetricDim()
 	r, _ := x.Dims()
 	if r != n {
 		panic(ErrShape)
@@ -495,7 +495,7 @@ func (s *SymDense) RankTwo(a Symmetric, alpha float64, x, y Vector) {
 
 // ScaleSym multiplies the elements of a by f, placing the result in the receiver.
 func (s *SymDense) ScaleSym(f float64, a Symmetric) {
-	n := a.Symmetric()
+	n := a.SymmetricDim()
 	s.reuseAsNonZeroed(n)
 	if a, ok := a.(RawSymmetricer); ok {
 		amat := a.RawSymmetric()
@@ -523,7 +523,7 @@ func (s *SymDense) ScaleSym(f float64, a Symmetric) {
 // have to be a strict subset, dimension repeats are allowed.
 func (s *SymDense) SubsetSym(a Symmetric, set []int) {
 	n := len(set)
-	na := a.Symmetric()
+	na := a.SymmetricDim()
 	s.reuseAsNonZeroed(n)
 	var restore func()
 	if a == s {
@@ -663,7 +663,7 @@ func (s *SymDense) GrowSym(n int) Symmetric {
 // PowPSD returns an error if the matrix is not positive symmetric definite
 // or the Eigen decomposition is not successful.
 func (s *SymDense) PowPSD(a Symmetric, pow float64) error {
-	dim := a.Symmetric()
+	dim := a.SymmetricDim()
 	s.reuseAsNonZeroed(dim)
 
 	var eigen EigenSym

mat/symmetric_test.go

@@ -659,7 +659,7 @@ func TestViewGrowSquare(t *testing.T) {
 
 		// Grow the matrix back to the original view
 		gn := n - start1 - start2
-		g := v2.GrowSym(gn - v2.Symmetric()).(*SymDense)
+		g := v2.GrowSym(gn - v2.SymmetricDim()).(*SymDense)
 		g.SetSym(1, 1, 2.2)
 
 		for i := 0; i < gn; i++ {

mat/triangular.go

@@ -659,7 +659,7 @@ func (t *TriDense) Trace() float64 {
 // copySymIntoTriangle copies a symmetric matrix into a TriDense
 func copySymIntoTriangle(t *TriDense, s Symmetric) {
 	n, upper := t.Triangle()
-	ns := s.Symmetric()
+	ns := s.SymmetricDim()
 	if n != ns {
 		panic("mat: triangle size mismatch")
 	}

optimize/cmaes.go

@@ -206,7 +206,7 @@ func (cma *CmaEsChol) Init(dim, tasks int) int {
 	cma.mean = resize(cma.mean, dim) // mean location initialized at the start of Run
 
 	if cma.InitCholesky != nil {
-		if cma.InitCholesky.Symmetric() != dim {
+		if cma.InitCholesky.SymmetricDim() != dim {
 			panic("cma-es-chol: incorrect InitCholesky size")
 		}
 		cma.chol.Clone(cma.InitCholesky)

optimize/functions/functions.go

@@ -60,7 +60,7 @@ func (Beale) Hess(dst *mat.SymDense, x []float64) {
 	if len(x) != 2 {
 		panic("dimension of the problem must be 2")
 	}
-	if len(x) != dst.Symmetric() {
+	if len(x) != dst.SymmetricDim() {
 		panic("incorrect size of the Hessian")
 	}
 
@@ -562,7 +562,7 @@ func (BrownBadlyScaled) Hess(dst *mat.SymDense, x []float64) {
 	if len(x) != 2 {
 		panic("dimension of the problem must be 2")
 	}
-	if len(x) != dst.Symmetric() {
+	if len(x) != dst.SymmetricDim() {
 		panic("incorrect size of the Hessian")
 	}
 
@@ -639,7 +639,7 @@ func (BrownAndDennis) Hess(dst *mat.SymDense, x []float64) {
 	if len(x) != 4 {
 		panic("dimension of the problem must be 4")
 	}
-	if len(x) != dst.Symmetric() {
+	if len(x) != dst.SymmetricDim() {
 		panic("incorrect size of the Hessian")
 	}
 
@@ -1274,7 +1274,7 @@ func (PowellBadlyScaled) Hess(dst *mat.SymDense, x []float64) {
 	if len(x) != 2 {
 		panic("dimension of the problem must be 2")
 	}
-	if len(x) != dst.Symmetric() {
+	if len(x) != dst.SymmetricDim() {
 		panic("incorrect size of the Hessian")
 	}
 
@@ -1522,7 +1522,7 @@ func (Watson) Grad(grad, x []float64) {
 
 func (Watson) Hess(dst *mat.SymDense, x []float64) {
 	dim := len(x)
-	if len(x) != dst.Symmetric() {
+	if len(x) != dst.SymmetricDim() {
 		panic("incorrect size of the Hessian")
 	}
 
@@ -1644,7 +1644,7 @@ func (Wood) Hess(dst *mat.SymDense, x []float64) {
 	if len(x) != 4 {
 		panic("dimension of the problem must be 4")
 	}
-	if len(x) != dst.Symmetric() {
+	if len(x) != dst.SymmetricDim() {
 		panic("incorrect size of the Hessian")
 	}
 

optimize/minimize.go

@@ -447,7 +447,7 @@ func getInitLocation(dim int, initX []float64, initValues *Location) (Operation,
 		op |= GradEvaluation
 	}
 	if initValues.Hessian != nil {
-		if initValues.Hessian.Symmetric() != dim {
+		if initValues.Hessian.SymmetricDim() != dim {
 			panic("optimize: initial Hessian does not match problem dimension")
 		}
 		loc.Hessian = initValues.Hessian

stat/distmat/wishart.go

@@ -44,7 +44,7 @@ type Wishart struct {
 //
 // NewWishart panics if nu <= d - 1 where d is the order of v.
 func NewWishart(v mat.Symmetric, nu float64, src rand.Source) (*Wishart, bool) {
-	dim := v.Symmetric()
+	dim := v.SymmetricDim()
 	if nu <= float64(dim-1) {
 		panic("wishart: nu must be greater than dim-1")
 	}
@@ -75,7 +75,7 @@ func NewWishart(v mat.Symmetric, nu float64, src rand.Source) (*Wishart, bool) {
 func (w *Wishart) MeanSymTo(dst *mat.SymDense) {
 	if dst.IsEmpty() {
 		dst.ReuseAsSym(w.dim)
-	} else if dst.Symmetric() != w.dim {
+	} else if dst.SymmetricDim() != w.dim {
 		panic(badDim)
 	}
 	w.setV()
@@ -94,7 +94,7 @@ func (w *Wishart) ProbSym(x mat.Symmetric) float64 {
 // LogProbSym returns -∞ if the input matrix is not positive definite (the Cholesky
 // decomposition fails).
 func (w *Wishart) LogProbSym(x mat.Symmetric) float64 {
-	dim := x.Symmetric()
+	dim := x.SymmetricDim()
 	if dim != w.dim {
 		panic(badDim)
 	}
@@ -109,7 +109,7 @@ func (w *Wishart) LogProbSym(x mat.Symmetric) float64 {
 // LogProbSymChol returns the log of the probability of the input symmetric matrix
 // given its Cholesky decomposition.
 func (w *Wishart) LogProbSymChol(cholX *mat.Cholesky) float64 {
-	dim := cholX.Symmetric()
+	dim := cholX.SymmetricDim()
 	if dim != w.dim {
 		panic(badDim)
 	}

stat/distmat/wishart_test.go

@@ -111,7 +111,7 @@ func TestWishartRand(t *testing.T) {
 		},
 	} {
 		rnd := rand.New(rand.NewSource(1))
-		dim := test.v.Symmetric()
+		dim := test.v.SymmetricDim()
 		w, ok := NewWishart(test.v, test.nu, rnd)
 		if !ok {
 			panic("bad test")

stat/distmv/dirichlet.go

@@ -66,7 +66,7 @@ func NewDirichlet(alpha []float64, src rand.Source) *Dirichlet {
 func (d *Dirichlet) CovarianceMatrix(dst *mat.SymDense) {
 	if dst.IsEmpty() {
 		*dst = *(dst.GrowSym(d.dim).(*mat.SymDense))
-	} else if dst.Symmetric() != d.dim {
+	} else if dst.SymmetricDim() != d.dim {
 		panic("dirichelet: input matrix size mismatch")
 	}
 	scale := 1 / (d.sumAlpha * d.sumAlpha * (d.sumAlpha + 1))

stat/distmv/general_test.go

@@ -77,7 +77,7 @@ type Cover interface {
 func checkCov(t *testing.T, cas int, x *mat.Dense, c Cover, tol float64) {
 	var cov mat.SymDense
 	c.CovarianceMatrix(&cov)
-	n := cov.Symmetric()
+	n := cov.SymmetricDim()
 	cov2 := mat.NewSymDense(n, nil)
 	c.CovarianceMatrix(cov2)
 	if !mat.Equal(&cov, cov2) {

stat/distmv/normal.go

@@ -43,7 +43,7 @@ func NewNormal(mu []float64, sigma mat.Symmetric, src rand.Source) (*Normal, boo
 	if len(mu) == 0 {
 		panic(badZeroDimension)
 	}
-	dim := sigma.Symmetric()
+	dim := sigma.SymmetricDim()
 	if dim != len(mu) {
 		panic(badSizeMismatch)
 	}
@@ -69,7 +69,7 @@ func NewNormal(mu []float64, sigma mat.Symmetric, src rand.Source) (*Normal, boo
 // panics if len(mu) is not equal to chol.Size().
 func NewNormalChol(mu []float64, chol *mat.Cholesky, src rand.Source) *Normal {
 	dim := len(mu)
-	if dim != chol.Symmetric() {
+	if dim != chol.SymmetricDim() {
 		panic(badSizeMismatch)
 	}
 	n := &Normal{
@@ -93,7 +93,7 @@ func NewNormalPrecision(mu []float64, prec *mat.SymDense, src rand.Source) (norm
 	if len(mu) == 0 {
 		panic(badZeroDimension)
 	}
-	dim := prec.Symmetric()
+	dim := prec.SymmetricDim()
 	if dim != len(mu) {
 		panic(badSizeMismatch)
 	}
@@ -161,7 +161,7 @@ func (n *Normal) ConditionNormal(observed []int, values []float64, src rand.Sour
 func (n *Normal) CovarianceMatrix(dst *mat.SymDense) {
 	if dst.IsEmpty() {
 		*dst = *(dst.GrowSym(n.dim).(*mat.SymDense))
-	} else if dst.Symmetric() != n.dim {
+	} else if dst.SymmetricDim() != n.dim {
 		panic("normal: input matrix size mismatch")
 	}
 	dst.CopySym(&n.sigma)
@@ -197,7 +197,7 @@ func NormalLogProb(x, mu []float64, chol *mat.Cholesky) float64 {
 	if len(x) != dim {
 		panic(badSizeMismatch)
 	}
-	if chol.Symmetric() != dim {
+	if chol.SymmetricDim() != dim {
 		panic(badSizeMismatch)
 	}
 	logSqrtDet := 0.5 * chol.LogDet()
@@ -301,7 +301,7 @@ func (n *Normal) Rand(x []float64) []float64 {
 // available. Otherwise, the NewNormal function should be used.
 func NormalRand(x, mean []float64, chol *mat.Cholesky, src rand.Source) []float64 {
 	x = reuseAs(x, len(mean))
-	if len(mean) != chol.Symmetric() {
+	if len(mean) != chol.SymmetricDim() {
 		panic(badInputLength)
 	}
 	if src == nil {

stat/distmv/normal_test.go

@@ -399,7 +399,7 @@ func TestCovarianceMatrix(t *testing.T) {
 		if !mat.EqualApprox(&cov, test.sigma, 1e-14) {
 			t.Errorf("Covariance mismatch with nil input")
 		}
-		dim := test.sigma.Symmetric()
+		dim := test.sigma.SymmetricDim()
 		cov = *mat.NewSymDense(dim, nil)
 		normal.CovarianceMatrix(&cov)
 		if !mat.EqualApprox(&cov, test.sigma, 1e-14) {

stat/distmv/studentst.go

@@ -55,7 +55,7 @@ func NewStudentsT(mu []float64, sigma mat.Symmetric, nu float64, src rand.Source
 	if len(mu) == 0 {
 		panic(badZeroDimension)
 	}
-	dim := sigma.Symmetric()
+	dim := sigma.SymmetricDim()
 	if dim != len(mu) {
 		panic(badSizeMismatch)
 	}
@@ -231,7 +231,7 @@ func findUnob(observed []int, dim int) (unobserved []int) {
 func (st *StudentsT) CovarianceMatrix(dst *mat.SymDense) {
 	if dst.IsEmpty() {
 		*dst = *(dst.GrowSym(st.dim).(*mat.SymDense))
-	} else if dst.Symmetric() != st.dim {
+	} else if dst.SymmetricDim() != st.dim {
 		panic("studentst: input matrix size mismatch")
 	}
 	dst.CopySym(&st.sigma)

stat/mds/mds.go

@@ -24,7 +24,7 @@ import (
 func TorgersonScaling(dst *mat.Dense, eigdst []float64, dis mat.Symmetric) (k int, eig []float64) {
 	// https://doi.org/10.1007/0-387-28981-X_12
 
-	n := dis.Symmetric()
+	n := dis.SymmetricDim()
 	if dst.IsEmpty() {
 		dst.ReuseAs(n, n)
 	} else {

stat/samplemv/metropolishastings.go

@@ -173,7 +173,7 @@ type ProposalNormal struct {
 //
 // NewProposalNormal returns {nil, false} if the covariance matrix is not positive-definite.
 func NewProposalNormal(sigma *mat.SymDense, src rand.Source) (*ProposalNormal, bool) {
-	mu := make([]float64, sigma.Symmetric())
+	mu := make([]float64, sigma.SymmetricDim())
 	normal, ok := distmv.NewNormal(mu, sigma, src)
 	if !ok {
 		return nil, false

stat/statmat.go

@@ -29,7 +29,7 @@ func CovarianceMatrix(dst *mat.SymDense, x mat.Matrix, weights []float64) {
 
 	if dst.IsEmpty() {
 		*dst = *(dst.GrowSym(c).(*mat.SymDense))
-	} else if n := dst.Symmetric(); n != c {
+	} else if n := dst.SymmetricDim(); n != c {
 		panic(mat.ErrShape)
 	}
 
@@ -86,7 +86,7 @@ func CorrelationMatrix(dst *mat.SymDense, x mat.Matrix, weights []float64) {
 
 // covToCorr converts a covariance matrix to a correlation matrix.
 func covToCorr(c *mat.SymDense) {
-	r := c.Symmetric()
+	r := c.SymmetricDim()
 
 	s := make([]float64, r)
 	for i := 0; i < r; i++ {

stat/statmat_test.go

@@ -244,7 +244,7 @@ func TestCorrCov(t *testing.T) {
 		CorrelationMatrix(&corr, test.data, test.weights)
 		CovarianceMatrix(&cov, test.data, test.weights)
 
-		r := cov.Symmetric()
+		r := cov.SymmetricDim()
 
 		// Get the diagonal elements from cov to determine the sigmas.
 		sigmas := make([]float64, r)
@@ -252,11 +252,11 @@ func TestCorrCov(t *testing.T) {
 			sigmas[i] = math.Sqrt(cov.At(i, i))
 		}
 
-		covFromCorr := mat.NewSymDense(corr.Symmetric(), nil)
+		covFromCorr := mat.NewSymDense(corr.SymmetricDim(), nil)
 		covFromCorr.CopySym(&corr)
 		corrToCov(covFromCorr, sigmas)
 
-		corrFromCov := mat.NewSymDense(cov.Symmetric(), nil)
+		corrFromCov := mat.NewSymDense(cov.SymmetricDim(), nil)
 		corrFromCov.CopySym(&cov)
 		covToCorr(corrFromCov)
 
@@ -450,7 +450,7 @@ func BenchmarkCovToCorr(b *testing.B) {
 	m := randMat(small, small)
 	var cov mat.SymDense
 	CovarianceMatrix(&cov, m, nil)
-	cc := mat.NewSymDense(cov.Symmetric(), nil)
+	cc := mat.NewSymDense(cov.SymmetricDim(), nil)
 	b.ResetTimer()
 	for i := 0; i < b.N; i++ {
 		b.StopTimer()
@@ -465,7 +465,7 @@ func BenchmarkCorrToCov(b *testing.B) {
 	m := randMat(small, small)
 	var corr mat.SymDense
 	CorrelationMatrix(&corr, m, nil)
-	cc := mat.NewSymDense(corr.Symmetric(), nil)
+	cc := mat.NewSymDense(corr.SymmetricDim(), nil)
 	sigma := make([]float64, small)
 	for i := range sigma {
 		sigma[i] = 2