Skip to content

Commit

Permalink
Towards #12251 and #12153, reimplement all full(X) method defs. as co…
Browse files Browse the repository at this point in the history 
…nvert(Array, X) and add convert(AbstractArray, X) methods for Factorizations (#17066)

* Implement convert(Matrix, SparseMatrixCSC) <- convert(Array, SparseMatrixCSC) chain, and reimplement full(SparseMatrixCSC) as a synonymous child of convert(Array, SparseMatrixCSC).

* Implement convert(Vector, AbstractSparseVector) <- convert(Array, AbstractSparseVector) chain, and reimplement full(AbstractSparseVector) as a synonymous child of convert(Array, AbstractSparseVector).

* Implement convert(Matrix, Bidiagonal) <- convert(Array, Bidiagonal) chain, and reimplement full(Bidiagonal) as a synonymous child of convert(Array, Bidiagonal).

* Implement convert(AbstractMatrix, X) <- convert(AbstractArray, X) <- convert(Matrix, X) <- convert(Array, X) chains for Cholesky and CholeskyPivoted types, and reimplement full(X) for both types as synonymous children of convert(Array, X).

* Implement convert(Matrix, Diagonal) <- convert(Array, Diagonal) chain, and reimplement full(Diagonal) as a synonymous child of convert(Array, Diagonal).

* Implement convert(AbstractMatrix, Eigen) <- convert(AbstractArray, Eigen) <- convert(Matrix, Eigen) <- convert(Array, Eigen) chain, and reimplement full(Eigen) as synonymous child of convert(Array, Eigen).

* Implement convert(AbstractMatrix, Hessenberg) <- convert(AbstractArray, Hessenberg) <- convert(Matrix, Hessenberg) <- convert(Array, Hessenberg) and convert(Matrix, HessenbergQ) <- convert(Array, HessenbergQ )chains, and reimplement full(X) for both types as synonmous children of convert(Array, X).

* Implement convert(SymTridiagonal, LDLt) <- convert(AbstractMatrix, LDLt) <- convert(AbstractArray, LDLt) <- convert(Matrix, LDLt) <- convert(Array, LDLt) chain, and reimplement full(LDLt) as a synonymous child of convert(Array, LDLt).

* Implement convert(AbstractMatrix, LQ) <- convert(AbstractArray, LQ) <- convert(Matrix, LQ) <- convert(Array, LQ) and convert(Matrix, LQPackedQ) <- convert(Array, LQPackedQ) chains, and reimplement full(X) for both types as a synonymous child of convert(Array, X).

* Implement convert(AbstractMatrix, X) < convert(AbstractArray, X) <- convert(Matrix, X) <- convert(Array, X) chains for LU and LU{T,Tridiagonal{T}} types, and reimplement full(X) for both types as synonymous children of convert(Array, X).

* Implement convert(AbstractArray, Union{QR,QRCompactWY}) <- convert(AbstractMatrix, Union{QR,QRCompactWY}) <- convert(Matrix, Union{QR,QRCompactWY}) <- convert(Array, Union{QR,QRCompactWY}) and convert(Matrix, Union{QRPackedQ,QRCompactWYQ}) <- convert(Array, Union{QRPackedQ,QRCompactWYQ}) chains, and reimplement full(Union{QR,QRCompactWY}) and full(Union{QRPackedQ,QRCompactWYQ}) as synonymous children of convert(Array, Union{QR,QRCompactWY}) and convert(Array, Union{QRPackedQ,QRCompactWYQ}) respectively.

* Implement convert(AbstractMatrix, Schur) <- convert(AbstractArray, Schur) <- convert(Matrix, Schur) <- convert(Array, Schur) chain, and reimplement full(Schur) as a synonymous child of convert(Array, Schur).

* Implement convert(AbstractMatrix, SVD) <- convert(AbstractArray, SVD) <- convert(Matrix, SVD) <- convert(Array, SVD) chain, and reimplement full(SVD) as a synonymous child of convert(Array, SVD).

* Implements convert(Matrix, Symmetric) and convert(Matrix, Hermitian). Implements convert(Array, Union{Symmetric,Hermitian}) as a short child of the two preceding methods, and reimplements full(Union{Symmetric,Hermitian}) likewise.

* Implements convert(Array, AbstractTriangular) as a short child of convert(Matrix, AbstractTriangular), and reimplements full(AbstractTriangular) as a short child of convert(Array, AbstractTriangular).

* Implements convert(Array, Tridiagonal) and convert(Array, SymTridiagonal) as synonymous children of convert(Matrix, Tridiagonal) and convert(Matrix, SymTridiagonal) respectively, and reimplements full(Tridiagonal) and full(SymTridiagonal) as short children of convert(Array, Tridiagonal) and convert(Array, SymTridiagonal) respectively.
  • Loading branch information
@Sacha0 @andreasnoack
Sacha0 authored and andreasnoack committed Jul 10, 2016
1 parent 4a84310 commit f8d67f7
Show file tree
Hide file tree
Showing 17 changed files with 179 additions and 104 deletions.
3 changes: 2 additions & 1 deletion base/linalg/bidiag.jl
View file
Original file line number Diff line number Diff line change
Expand Up @@ -114,7 +114,6 @@ function Base.replace_in_print_matrix(A::Bidiagonal,i::Integer,j::Integer,s::Abs
end

#Converting from Bidiagonal to dense Matrix
full{T}(M::Bidiagonal{T}) = convert(Matrix{T}, M)
function convert{T}(::Type{Matrix{T}}, A::Bidiagonal)
n = size(A, 1)
B = zeros(T, n, n)
Expand All @@ -130,6 +129,8 @@ function convert{T}(::Type{Matrix{T}}, A::Bidiagonal)
return B
end
convert{T}(::Type{Matrix}, A::Bidiagonal{T}) = convert(Matrix{T}, A)
convert(::Type{Array}, A::Bidiagonal) = convert(Matrix, A)
full(A::Bidiagonal) = convert(Array, A)
promote_rule{T,S}(::Type{Matrix{T}}, ::Type{Bidiagonal{S}})=Matrix{promote_type(T,S)}

#Converting from Bidiagonal to Tridiagonal
Expand Down
15 changes: 12 additions & 3 deletions base/linalg/cholesky.jl
View file
Original file line number Diff line number Diff line change
Expand Up @@ -324,11 +324,20 @@ convert{T}(::Type{CholeskyPivoted{T}},C::CholeskyPivoted) =
CholeskyPivoted(AbstractMatrix{T}(C.factors),C.uplo,C.piv,C.rank,C.tol,C.info)
convert{T}(::Type{Factorization{T}}, C::CholeskyPivoted) = convert(CholeskyPivoted{T}, C)

full{T,S}(C::Cholesky{T,S}) = C.uplo == 'U' ? C[:U]'C[:U] : C[:L]*C[:L]'
function full(F::CholeskyPivoted)
convert(::Type{AbstractMatrix}, C::Cholesky) = C.uplo == 'U' ? C[:U]'C[:U] : C[:L]*C[:L]'
convert(::Type{AbstractArray}, C::Cholesky) = convert(AbstractMatrix, C)
convert(::Type{Matrix}, C::Cholesky) = convert(Array, convert(AbstractArray, C))
convert(::Type{Array}, C::Cholesky) = convert(Matrix, C)
full(C::Cholesky) = convert(Array, C)

function convert(::Type{AbstractMatrix}, F::CholeskyPivoted)
ip = invperm(F[:p])
return (F[:L] * F[:U])[ip,ip]
(F[:L] * F[:U])[ip,ip]
end
convert(::Type{AbstractArray}, F::CholeskyPivoted) = convert(AbstractMatrix, F)
convert(::Type{Matrix}, F::CholeskyPivoted) = convert(Array, convert(AbstractArray, F))
convert(::Type{Array}, F::CholeskyPivoted) = convert(Matrix, F)
full(F::CholeskyPivoted) = convert(Array, F)

copy(C::Cholesky) = Cholesky(copy(C.factors), C.uplo)
copy(C::CholeskyPivoted) = CholeskyPivoted(copy(C.factors), C.uplo, C.piv, C.rank, C.tol, C.info)
Expand Down
5 changes: 3 additions & 2 deletions base/linalg/diagonal.jl
View file
Original file line number Diff line number Diff line change
Expand Up @@ -23,6 +23,9 @@ convert{T}(::Type{Diagonal{T}}, D::Diagonal) = Diagonal{T}(convert(Vector{T}, D.
convert{T}(::Type{AbstractMatrix{T}}, D::Diagonal) = convert(Diagonal{T}, D)
convert{T}(::Type{UpperTriangular}, A::Diagonal{T}) = UpperTriangular(A)
convert{T}(::Type{LowerTriangular}, A::Diagonal{T}) = LowerTriangular(A)
convert(::Type{Matrix}, D::Diagonal) = diagm(D.diag)
convert(::Type{Array}, D::Diagonal) = convert(Matrix, D)
full(D::Diagonal) = convert(Array, D)

function similar{T}(D::Diagonal, ::Type{T})
return Diagonal{T}(similar(D.diag, T))
Expand All @@ -39,8 +42,6 @@ function size(D::Diagonal,d::Integer)
return d<=2 ? length(D.diag) : 1
end

full(D::Diagonal) = diagm(D.diag)

@inline function getindex(D::Diagonal, i::Int, j::Int)
@boundscheck checkbounds(D, i, j)
if i == j
Expand Down
8 changes: 7 additions & 1 deletion base/linalg/eigen.jl
View file
Original file line number Diff line number Diff line change
Expand Up @@ -182,5 +182,11 @@ end

eigvecs(A::AbstractMatrix, B::AbstractMatrix) = eigvecs(eigfact(A, B))

# Conversion methods

## Can we determine the source/result is Real? This is not stored in the type Eigen
full(F::Eigen) = F.vectors * Diagonal(F.values) / F.vectors
convert(::Type{AbstractMatrix}, F::Eigen) = F.vectors * Diagonal(F.values) / F.vectors
convert(::Type{AbstractArray}, F::Eigen) = convert(AbstractMatrix, F)
convert(::Type{Matrix}, F::Eigen) = convert(Array, convert(AbstractArray, F))
convert(::Type{Array}, F::Eigen) = convert(Matrix, F)
full(F::Eigen) = convert(Array, F)
13 changes: 8 additions & 5 deletions base/linalg/hessenberg.jl
View file
Original file line number Diff line number Diff line change
Expand Up @@ -53,11 +53,14 @@ function getindex(A::HessenbergQ, i::Integer, j::Integer)
end

## reconstruct the original matrix
full{T<:BlasFloat}(A::HessenbergQ{T}) = LAPACK.orghr!(1, size(A.factors, 1), copy(A.factors), A.τ)
function full(F::Hessenberg)
fq = full(F[:Q])
return (fq * F[:H]) * fq'
end
convert{T<:BlasFloat}(::Type{Matrix}, A::HessenbergQ{T}) = LAPACK.orghr!(1, size(A.factors, 1), copy(A.factors), A.τ)
convert(::Type{Array}, A::HessenbergQ) = convert(Matrix, A)
full(A::HessenbergQ) = convert(Array, A)
convert(::Type{AbstractMatrix}, F::Hessenberg) = (fq = full(F[:Q]); (fq * F[:H]) * fq')
convert(::Type{AbstractArray}, F::Hessenberg) = convert(AbstractMatrix, F)
convert(::Type{Matrix}, F::Hessenberg) = convert(Array, convert(AbstractArray, F))
convert(::Type{Array}, F::Hessenberg) = convert(Matrix, F)
full(F::Hessenberg) = convert(Array, F)

A_mul_B!{T<:BlasFloat}(Q::HessenbergQ{T}, X::StridedVecOrMat{T}) =
LAPACK.ormhr!('L', 'N', 1, size(Q.factors, 1), Q.factors, Q.τ, X)
Expand Down
9 changes: 7 additions & 2 deletions base/linalg/ldlt.jl
View file
Original file line number Diff line number Diff line change
Expand Up @@ -76,11 +76,16 @@ function A_ldiv_B!{T}(S::LDLt{T,SymTridiagonal{T}}, B::AbstractVecOrMat{T})
return B
end

## reconstruct the original matrix, which is tridiagonal
function full(F::LDLt)
# Conversion methods
function convert(::Type{SymTridiagonal}, F::LDLt)
e = copy(F.data.ev)
d = copy(F.data.dv)
e .*= d[1:end-1]
d[2:end] += e .* F.data.ev
SymTridiagonal(d, e)
end
convert(::Type{AbstractMatrix}, F::LDLt) = convert(SymTridiagonal, F)
convert(::Type{AbstractArray}, F::LDLt) = convert(AbstractMatrix, F)
convert(::Type{Matrix}, F::LDLt) = convert(Array, convert(AbstractArray, F))
convert(::Type{Array}, F::LDLt) = convert(Matrix, F)
full(F::LDLt) = convert(Array, F)
39 changes: 22 additions & 17 deletions base/linalg/lq.jl
View file
Original file line number Diff line number Diff line change
Expand Up @@ -46,8 +46,15 @@ function lq(A::Union{Number, AbstractMatrix}; thin::Bool=true)
end

copy(A::LQ) = LQ(copy(A.factors), copy(A.τ))

convert{T}(::Type{LQ{T}},A::LQ) = LQ(convert(AbstractMatrix{T}, A.factors), convert(Vector{T}, A.τ))
convert{T}(::Type{Factorization{T}}, A::LQ) = convert(LQ{T}, A)
convert(::Type{AbstractMatrix}, A::LQ) = A[:L]*A[:Q]
convert(::Type{AbstractArray}, A::LQ) = convert(AbstractMatrix, A)
convert(::Type{Matrix}, A::LQ) = convert(Array, convert(AbstractArray, A))
convert(::Type{Array}, A::LQ) = convert(Matrix, A)
full(A::LQ) = convert(Array, A)

ctranspose{T}(A::LQ{T}) = QR{T,typeof(A.factors)}(A.factors', A.τ)

function getindex(A::LQ, d::Symbol)
Expand All @@ -73,6 +80,21 @@ getq(A::LQ) = LQPackedQ(A.factors, A.τ)

convert{T}(::Type{LQPackedQ{T}}, Q::LQPackedQ) = LQPackedQ(convert(AbstractMatrix{T}, Q.factors), convert(Vector{T}, Q.τ))
convert{T}(::Type{AbstractMatrix{T}}, Q::LQPackedQ) = convert(LQPackedQ{T}, Q)
convert(::Type{Matrix}, A::LQPackedQ) = LAPACK.orglq!(copy(A.factors),A.τ)
convert(::Type{Array}, A::LQPackedQ) = convert(Matrix, A)
function full{T}(A::LQPackedQ{T}; thin::Bool = true)
#= We construct the full eye here, even though it seems inefficient, because
every element in the output matrix is a function of all the elements of
the input matrix. The eye is modified by the elementary reflectors held
in A, so this is not just an indexing operation. Note that in general
explicitly constructing Q, rather than using the ldiv or mult methods,
may be a wasteful allocation. =#
if thin
convert(Array, A)
else
A_mul_B!(A, eye(T, size(A.factors,2), size(A.factors,1)))
end
end

size(A::LQ, dim::Integer) = size(A.factors, dim)
size(A::LQ) = size(A.factors)
Expand All @@ -88,23 +110,6 @@ end

size(A::LQPackedQ) = size(A.factors)

full(A::LQ) = A[:L]*A[:Q]
#=
We construct the full eye here, even though it seems ineffecient, because
every element in the output matrix is a function of all the elements of
the input matrix. The eye is modified by the elementary reflectors held
in A, so this is not just an indexing operation. Note that in general
explicitly constructing Q, rather than using the ldiv or mult methods,
may be a wasteful allocation.
=#
function full{T}(A::LQPackedQ{T}; thin::Bool=true)
if thin
LAPACK.orglq!(copy(A.factors),A.τ)
else
A_mul_B!(A, eye(T, size(A.factors,2), size(A.factors,1)))
end
end

## Multiplication by LQ
A_mul_B!{T<:BlasFloat}(A::LQ{T}, B::StridedVecOrMat{T}) = A[:L]*LAPACK.ormlq!('L','N',A.factors,A.τ,B)
A_mul_B!{T<:BlasFloat}(A::LQ{T}, B::QR{T}) = A[:L]*LAPACK.ormlq!('L','N',A.factors,A.τ,full(B))
Expand Down
83 changes: 46 additions & 37 deletions base/linalg/lu.jl
View file
Original file line number Diff line number Diff line change
Expand Up @@ -329,41 +329,6 @@ function getindex{T}(F::Base.LinAlg.LU{T,Tridiagonal{T}}, d::Symbol)
throw(KeyError(d))
end

function full{T}(F::Base.LinAlg.LU{T,Tridiagonal{T}})
n = size(F, 1)

dl = copy(F.factors.dl)
d = copy(F.factors.d)
du = copy(F.factors.du)
du2 = copy(F.factors.du2)

for i = n - 1:-1:1
li = dl[i]
dl[i] = li*d[i]
d[i + 1] += li*du[i]
if i < n - 1
du[i + 1] += li*du2[i]
end

if F.ipiv[i] != i
tmp = dl[i]
dl[i] = d[i]
d[i] = tmp

tmp = d[i + 1]
d[i + 1] = du[i]
du[i] = tmp

if i < n - 1
tmp = du[i + 1]
du[i + 1] = du2[i]
du2[i] = tmp
end
end
end
return Tridiagonal(dl, d, du)
end

# See dgtts2.f
function A_ldiv_B!{T}(A::LU{T,Tridiagonal{T}}, B::AbstractVecOrMat)
n = size(A,1)
Expand Down Expand Up @@ -467,5 +432,49 @@ end

/(B::AbstractMatrix,A::LU) = At_ldiv_Bt(A,B).'

## reconstruct the original matrix
full(F::LU) = (F[:L] * F[:U])[invperm(F[:p]),:]
# Conversions
convert(::Type{AbstractMatrix}, F::LU) = (F[:L] * F[:U])[invperm(F[:p]),:]
convert(::Type{AbstractArray}, F::LU) = convert(AbstractMatrix, F)
convert(::Type{Matrix}, F::LU) = convert(Array, convert(AbstractArray, F))
convert(::Type{Array}, F::LU) = convert(Matrix, F)
full(F::LU) = convert(Array, F)

function convert{T}(::Type{Tridiagonal}, F::Base.LinAlg.LU{T,Tridiagonal{T}})
n = size(F, 1)

dl = copy(F.factors.dl)
d = copy(F.factors.d)
du = copy(F.factors.du)
du2 = copy(F.factors.du2)

for i = n - 1:-1:1
li = dl[i]
dl[i] = li*d[i]
d[i + 1] += li*du[i]
if i < n - 1
du[i + 1] += li*du2[i]
end

if F.ipiv[i] != i
tmp = dl[i]
dl[i] = d[i]
d[i] = tmp

tmp = d[i + 1]
d[i + 1] = du[i]
du[i] = tmp

if i < n - 1
tmp = du[i + 1]
du[i + 1] = du2[i]
du2[i] = tmp
end
end
end
return Tridiagonal(dl, d, du)
end
convert{T}(::Type{AbstractMatrix}, F::Base.LinAlg.LU{T,Tridiagonal{T}}) = convert(Tridiagonal, F)
convert{T}(::Type{AbstractArray}, F::Base.LinAlg.LU{T,Tridiagonal{T}}) = convert(AbstractMatrix, F)
convert{T}(::Type{Matrix}, F::Base.LinAlg.LU{T,Tridiagonal{T}}) = convert(Array, convert(AbstractArray, F))
convert{T}(::Type{Array}, F::Base.LinAlg.LU{T,Tridiagonal{T}}) = convert(Matrix, F)
full{T}(F::Base.LinAlg.LU{T,Tridiagonal{T}}) = convert(Array, F)
33 changes: 23 additions & 10 deletions base/linalg/qr.jl
View file
Original file line number Diff line number Diff line change
Expand Up @@ -234,13 +234,23 @@ function qr!(v::AbstractVector)
__normalize!(v, nrm), nrm
end


convert{T}(::Type{QR{T}},A::QR) = QR(convert(AbstractMatrix{T}, A.factors), convert(Vector{T}, A.τ))
# Conversions
convert{T}(::Type{QR{T}}, A::QR) = QR(convert(AbstractMatrix{T}, A.factors), convert(Vector{T}, A.τ))
convert{T}(::Type{Factorization{T}}, A::QR) = convert(QR{T}, A)
convert{T}(::Type{QRCompactWY{T}},A::QRCompactWY) = QRCompactWY(convert(AbstractMatrix{T}, A.factors), convert(AbstractMatrix{T}, A.T))
convert{T}(::Type{QRCompactWY{T}}, A::QRCompactWY) = QRCompactWY(convert(AbstractMatrix{T}, A.factors), convert(AbstractMatrix{T}, A.T))
convert{T}(::Type{Factorization{T}}, A::QRCompactWY) = convert(QRCompactWY{T}, A)
convert{T}(::Type{QRPivoted{T}},A::QRPivoted) = QRPivoted(convert(AbstractMatrix{T}, A.factors), convert(Vector{T}, A.τ), A.jpvt)
convert(::Type{AbstractMatrix}, F::Union{QR,QRCompactWY}) = F[:Q] * F[:R]
convert(::Type{AbstractArray}, F::Union{QR,QRCompactWY}) = convert(AbstractMatrix, F)
convert(::Type{Matrix}, F::Union{QR,QRCompactWY}) = convert(Array, convert(AbstractArray, F))
convert(::Type{Array}, F::Union{QR,QRCompactWY}) = convert(Matrix, F)
full(F::Union{QR,QRCompactWY}) = convert(Array, F)
convert{T}(::Type{QRPivoted{T}}, A::QRPivoted) = QRPivoted(convert(AbstractMatrix{T}, A.factors), convert(Vector{T}, A.τ), A.jpvt)
convert{T}(::Type{Factorization{T}}, A::QRPivoted) = convert(QRPivoted{T}, A)
convert(::Type{AbstractMatrix}, F::QRPivoted) = (F[:Q] * F[:R])[:,invperm(F[:p])]
convert(::Type{AbstractArray}, F::QRPivoted) = convert(AbstractMatrix, F)
convert(::Type{Matrix}, F::QRPivoted) = convert(Array, convert(AbstractArray, F))
convert(::Type{Array}, F::QRPivoted) = convert(Matrix, F)
full(F::QRPivoted) = convert(Array, F)

function getindex(A::QR, d::Symbol)
m, n = size(A)
Expand Down Expand Up @@ -283,11 +293,6 @@ function getindex{T}(A::QRPivoted{T}, d::Symbol)
end
end

## reconstruct the original matrix
full(F::QR) = F[:Q] * F[:R]
full(F::QRCompactWY) = F[:Q] * F[:R]
full(F::QRPivoted) = (F[:Q] * F[:R])[:,invperm(F[:p])]

# Type-stable interface to get Q
getq(A::QRCompactWY) = QRCompactWYQ(A.factors,A.T)
getq(A::Union{QR, QRPivoted}) = QRPackedQ(A.factors,A.τ)
Expand All @@ -310,13 +315,21 @@ convert{T}(::Type{QRPackedQ{T}}, Q::QRPackedQ) = QRPackedQ(convert(AbstractMatri
convert{T}(::Type{AbstractMatrix{T}}, Q::QRPackedQ) = convert(QRPackedQ{T}, Q)
convert{S}(::Type{QRCompactWYQ{S}}, Q::QRCompactWYQ) = QRCompactWYQ(convert(AbstractMatrix{S}, Q.factors), convert(AbstractMatrix{S}, Q.T))
convert{S}(::Type{AbstractMatrix{S}}, Q::QRCompactWYQ) = convert(QRCompactWYQ{S}, Q)
convert{T}(::Type{Matrix}, A::Union{QRPackedQ{T},QRCompactWYQ{T}}) = A_mul_B!(A, eye(T, size(A.factors, 1), minimum(size(A.factors))))
convert(::Type{Array}, A::Union{QRPackedQ,QRCompactWYQ}) = convert(Matrix, A)
function full{T}(A::Union{QRPackedQ{T},QRCompactWYQ{T}}; thin::Bool = true)
if thin
convert(Array, A)
else
A_mul_B!(A, eye(T, size(A.factors, 1)))
end
end

size(A::Union{QR,QRCompactWY,QRPivoted}, dim::Integer) = size(A.factors, dim)
size(A::Union{QR,QRCompactWY,QRPivoted}) = size(A.factors)
size(A::Union{QRPackedQ,QRCompactWYQ}, dim::Integer) = 0 < dim ? (dim <= 2 ? size(A.factors, 1) : 1) : throw(BoundsError())
size(A::Union{QRPackedQ,QRCompactWYQ}) = size(A, 1), size(A, 2)

full{T}(A::Union{QRPackedQ{T},QRCompactWYQ{T}}; thin::Bool=true) = A_mul_B!(A, thin ? eye(T, size(A.factors,1), minimum(size(A.factors))) : eye(T, size(A.factors,1)))

function getindex(A::Union{QRPackedQ,QRCompactWYQ}, i::Integer, j::Integer)
x = zeros(eltype(A), size(A, 1))
Expand Down
7 changes: 6 additions & 1 deletion base/linalg/schur.jl
View file
Original file line number Diff line number Diff line change
Expand Up @@ -206,7 +206,12 @@ function schur(A::StridedMatrix, B::StridedMatrix)
SchurF[:S], SchurF[:T], SchurF[:Q], SchurF[:Z], SchurF[:alpha], SchurF[:beta]
end

full(F::Schur) = (F.Z * F.T) * F.Z'
# Conversion
convert(::Type{AbstractMatrix}, F::Schur) = (F.Z * F.T) * F.Z'
convert(::Type{AbstractArray}, F::Schur) = convert(AbstractMatrix, F)
convert(::Type{Matrix}, F::Schur) = convert(Array, convert(AbstractArray, F))
convert(::Type{Array}, F::Schur) = convert(Matrix, F)
full(F::Schur) = convert(Array, F)

copy(F::Schur) = Schur(copy(F.T), copy(F.Z), copy(F.values))
copy(F::GeneralizedSchur) = GeneralizedSchur(copy(F.S), copy(F.T), copy(F.alpha), copy(F.beta), copy(F.Q), copy(F.Z))
1 change: 0 additions & 1 deletion base/linalg/special.jl
View file
Original file line number Diff line number Diff line change
Expand Up @@ -8,7 +8,6 @@ convert{T}(::Type{SymTridiagonal}, A::Diagonal{T})=SymTridiagonal(A.diag, zeros(
convert{T}(::Type{Tridiagonal}, A::Diagonal{T})=Tridiagonal(zeros(T, size(A.diag,1)-1), A.diag, zeros(T, size(A.diag,1)-1))
convert(::Type{LowerTriangular}, A::Bidiagonal) = !A.isupper ? LowerTriangular(full(A)) : throw(ArgumentError("Bidiagonal matrix must have lower off diagonal to be converted to LowerTriangular"))
convert(::Type{UpperTriangular}, A::Bidiagonal) = A.isupper ? UpperTriangular(full(A)) : throw(ArgumentError("Bidiagonal matrix must have upper off diagonal to be converted to UpperTriangular"))
convert(::Type{Matrix}, D::Diagonal) = diagm(D.diag)

function convert(::Type{UnitUpperTriangular}, A::Diagonal)
if !all(A.diag .== one(eltype(A)))
Expand Down
7 changes: 6 additions & 1 deletion base/linalg/svd.jl
View file
Original file line number Diff line number Diff line change
Expand Up @@ -236,4 +236,9 @@ function svdvals{TA,TB}(A::StridedMatrix{TA}, B::StridedMatrix{TB})
return svdvals!(copy_oftype(A, S), copy_oftype(B, S))
end

full(F::SVD) = (F.U * Diagonal(F.S)) * F.Vt
# Conversion
convert(::Type{AbstractMatrix}, F::SVD) = (F.U * Diagonal(F.S)) * F.Vt
convert(::Type{AbstractArray}, F::SVD) = convert(AbstractMatrix, F)
convert(::Type{Matrix}, F::SVD) = convert(Array, convert(AbstractArray, F))
convert(::Type{Array}, F::SVD) = convert(Matrix, F)
full(F::SVD) = convert(Array, F)
8 changes: 6 additions & 2 deletions base/linalg/symmetric.jl
View file
Original file line number Diff line number Diff line change
Expand Up @@ -80,15 +80,19 @@ function similar{T}(A::Hermitian, ::Type{T})
return Hermitian(B)
end

full(A::Symmetric) = copytri!(copy(A.data), A.uplo)
full(A::Hermitian) = copytri!(copy(A.data), A.uplo, true)
# Conversion
convert(::Type{Matrix}, A::Symmetric) = copytri!(convert(Matrix, copy(A.data)), A.uplo)
convert(::Type{Matrix}, A::Hermitian) = copytri!(convert(Matrix, copy(A.data)), A.uplo, true)
convert(::Type{Array}, A::Union{Symmetric,Hermitian}) = convert(Matrix, A)
full(A::Union{Symmetric,Hermitian}) = convert(Array, A)
parent(A::HermOrSym) = A.data
convert{T,S<:AbstractMatrix}(::Type{Symmetric{T,S}},A::Symmetric{T,S}) = A
convert{T,S<:AbstractMatrix}(::Type{Symmetric{T,S}},A::Symmetric) = Symmetric{T,S}(convert(S,A.data),A.uplo)
convert{T}(::Type{AbstractMatrix{T}}, A::Symmetric) = Symmetric(convert(AbstractMatrix{T}, A.data), Symbol(A.uplo))
convert{T,S<:AbstractMatrix}(::Type{Hermitian{T,S}},A::Hermitian{T,S}) = A
convert{T,S<:AbstractMatrix}(::Type{Hermitian{T,S}},A::Hermitian) = Hermitian{T,S}(convert(S,A.data),A.uplo)
convert{T}(::Type{AbstractMatrix{T}}, A::Hermitian) = Hermitian(convert(AbstractMatrix{T}, A.data), Symbol(A.uplo))

copy{T,S}(A::Symmetric{T,S}) = (B = copy(A.data); Symmetric{T,typeof(B)}(B,A.uplo))
copy{T,S}(A::Hermitian{T,S}) = (B = copy(A.data); Hermitian{T,typeof(B)}(B,A.uplo))
ishermitian(A::Hermitian) = true
Expand Down
3 changes: 2 additions & 1 deletion base/linalg/triangular.jl
View file
Original file line number Diff line number Diff line change
Expand Up @@ -48,7 +48,8 @@ imag(A::LowerTriangular) = LowerTriangular(imag(A.data))
imag(A::UnitLowerTriangular) = LowerTriangular(tril!(imag(A.data),-1))
imag(A::UnitUpperTriangular) = UpperTriangular(triu!(imag(A.data),1))

full(A::AbstractTriangular) = convert(Matrix, A)
convert(::Type{Array}, A::AbstractTriangular) = convert(Matrix, A)
full(A::AbstractTriangular) = convert(Array, A)
parent(A::AbstractTriangular) = A.data

# then handle all methods that requires specific handling of upper/lower and unit diagonal
Expand Down
Loading

4 comments on commit f8d67f7

@nanosoldier
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Executing the daily benchmark build, I will reply here when finished:

@nanosoldier runbenchmarks(ALL, isdaily = true)

@nanosoldier
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Your benchmark job has completed - possible performance regressions were detected. A full report can be found here. cc @jrevels

@nanosoldier
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Executing the daily benchmark build, I will reply here when finished:

@nanosoldier runbenchmarks(ALL, isdaily = true)

@nanosoldier
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Your benchmark job has completed - possible performance regressions were detected. A full report can be found here. cc @jrevels

Please sign in to comment.