Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

stdlib: make dot product of Hermitian matrices real #52318

Merged
merged 6 commits into from
Feb 3, 2024
+5 −5
Merged

stdlib: make dot product of Hermitian matrices real #52318

Changes from 4 commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
525edb5
make dot product of Hermitian matrices real
araujoms Nov 27, 2023
9f1f5ea
restrict to real or complex Hermitian matrices
araujoms Nov 28, 2023
82cc8c5
Merge branch 'master' into patch-1
dkarrasch Dec 5, 2023
5097097
change into extensible version
araujoms Dec 7, 2023
58f5ed8
Revert "change into extensible version"
araujoms Dec 30, 2023
e7c27fb
Merge branch 'master' into patch-1
dkarrasch Feb 1, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
73 changes: 36 additions & 37 deletions stdlib/LinearAlgebra/src/symmetric.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -453,47 +453,46 @@ function triu(A::Symmetric, k::Integer=0)
end
end

for (T, trans, real) in [(:Symmetric, :transpose, :identity), (:(Hermitian{<:Union{Real,Complex}}), :adjoint, :real)]
@eval begin
function dot(A::$T, B::$T)
n = size(A, 2)
if n != size(B, 2)
throw(DimensionMismatch("A has dimensions $(size(A)) but B has dimensions $(size(B))"))
end
dot(A::Symmetric, B::Symmetric) = _dot_hermsym(A, B, transpose, identity)
dot(A::Hermitian{<:Union{Real,Complex}}, B::Hermitian{<:Union{Real,Complex}}) = _dot_hermsym(A, B, conj, real)

dotprod = zero(dot(first(A), first(B)))
@inbounds if A.uplo == 'U' && B.uplo == 'U'
for j in 1:n
for i in 1:(j - 1)
dotprod += 2 * $real(dot(A.data[i, j], B.data[i, j]))
end
dotprod += dot(A[j, j], B[j, j])
end
elseif A.uplo == 'L' && B.uplo == 'L'
for j in 1:n
dotprod += dot(A[j, j], B[j, j])
for i in (j + 1):n
dotprod += 2 * $real(dot(A.data[i, j], B.data[i, j]))
end
end
elseif A.uplo == 'U' && B.uplo == 'L'
for j in 1:n
for i in 1:(j - 1)
dotprod += 2 * $real(dot(A.data[i, j], $trans(B.data[j, i])))
end
dotprod += dot(A[j, j], B[j, j])
end
else
for j in 1:n
dotprod += dot(A[j, j], B[j, j])
for i in (j + 1):n
dotprod += 2 * $real(dot(A.data[i, j], $trans(B.data[j, i])))
end
end
function _dot_hermsym(A, B, trans, real)
n = size(A, 2)
if n != size(B, 2)
throw(DimensionMismatch("A has dimensions $(size(A)) but B has dimensions $(size(B))"))
end

dotprod = real(zero(dot(first(A), first(B))))
@inbounds if A.uplo == 'U' && B.uplo == 'U'
Copy link
Contributor

Choose a reason for hiding this comment

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

Can we stop using this? This function assumes but does not enforce 1-based indexing.

Copy link
Member

Choose a reason for hiding this comment

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

The parents of Symmetric and Hermitian are required to be one-based at construction (see the very top of this file), so this is safe.

for j in 1:n
for i in 1:(j-1)
dotprod += 2 * real(dot(A.data[i, j], B.data[i, j]))
end
dotprod += real(dot(A[j, j], B[j, j]))
end
elseif A.uplo == 'L' && B.uplo == 'L'
for j in 1:n
dotprod += real(dot(A[j, j], B[j, j]))
for i in (j+1):n
dotprod += 2 * real(dot(A.data[i, j], B.data[i, j]))
end
end
elseif A.uplo == 'U' && B.uplo == 'L'
for j in 1:n
for i in 1:(j-1)
dotprod += 2 * real(dot(A.data[i, j], trans(B.data[j, i])))
end
dotprod += real(dot(A[j, j], B[j, j]))
end
else
for j in 1:n
dotprod += real(dot(A[j, j], B[j, j]))
for i in (j+1):n
dotprod += 2 * real(dot(A.data[i, j], trans(B.data[j, i])))
end
return dotprod
end
end
return dotprod
end

(-)(A::Symmetric) = Symmetric(-A.data, sym_uplo(A.uplo))
Expand Down