Skip to content

Allows more generic matrix-vector products #140

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

Merged
merged 1 commit into from
Mar 20, 2020
Merged

Allows more generic matrix-vector products #140

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
10 changes: 5 additions & 5 deletions src/adjtrans.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -92,7 +92,7 @@ function *(op :: AdjointLinearOperator{T,S}, v :: AbstractVector{U}) where {T,S,
ishermitian(p) && return p * v
if p.ctprod !== nothing
increase_nctprod(p)
return p.ctprod(v)::Vector{promote_type(T,U)}
return p.ctprod(v)::typeof(v).name.wrapper{promote_type(T,U), typeof(v).parameters[2]}
Copy link
Member Author

@amontoison amontoison Mar 14, 2020 •
edited
Loading

Choose a reason for hiding this comment

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

typeof(v).parameters[2] represents the dimension of the Array if it's dense (1 for Vectors) or the type of Integer that is used for storing coefficients when it's sparse.

end
tprod = p.tprod
increment_tprod = true
Expand All @@ -109,7 +109,7 @@ function *(op :: AdjointLinearOperator{T,S}, v :: AbstractVector{U}) where {T,S,
else
increase_nprod(p)
end
return conj.(tprod(conj.(v)))::Vector{promote_type(T,U)}
return conj.(tprod(conj.(v)))::typeof(v).name.wrapper{promote_type(T,U), typeof(v).parameters[2]}
end

function *(op :: TransposeLinearOperator{T,S}, v :: AbstractVector{U}) where {T,S,U}
Expand All @@ -118,7 +118,7 @@ function *(op :: TransposeLinearOperator{T,S}, v :: AbstractVector{U}) where {T,
issymmetric(p) && return p * v
if p.tprod !== nothing
increase_ntprod(p)
return p.tprod(v)::Vector{promote_type(T,U)}
return p.tprod(v)::typeof(v).name.wrapper{promote_type(T,U), typeof(v).parameters[2]}
end
increment_ctprod = true
ctprod = p.ctprod
Expand All @@ -135,12 +135,12 @@ function *(op :: TransposeLinearOperator{T,S}, v :: AbstractVector{U}) where {T,
else
increase_nprod(p)
end
return conj.(ctprod(conj.(v)))::Vector{promote_type(T,U)}
return conj.(ctprod(conj.(v)))::typeof(v).name.wrapper{promote_type(T,U), typeof(v).parameters[2]}
end

function *(op :: ConjugateLinearOperator{T,S}, v :: AbstractVector{U}) where {T,S,U}
p = op.parent
return conj.(p * conj.(v))::Vector{promote_type(T,U)}
return conj.(p * conj.(v))::typeof(v).name.wrapper{promote_type(T,U), typeof(v).parameters[2]}
end

-(op :: AdjointLinearOperator) = adjoint(-op.parent)
Expand Down
2 changes: 1 addition & 1 deletion src/operations.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@ import Base.+, Base.-, Base.*, LinearAlgebra.mul!
function *(op :: AbstractLinearOperator{T}, v :: AbstractVector{S}) where {T,S}
size(v, 1) == size(op, 2) || throw(LinearOperatorException("shape mismatch"))
increase_nprod(op)
op.prod(v)::Vector{promote_type(T,S)}
op.prod(v)::typeof(v).name.wrapper{promote_type(T,S), typeof(v).parameters[2]}
end

# Unary operations.
Expand Down
10 changes: 10 additions & 0 deletions test/test_linop.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -646,6 +646,16 @@ function test_linop()
@test norm(Matrix(transpose(M)) - transpose(D)) ≤ sqrt(eps()) * norm(D)
@test norm(Matrix(M') - D') ≤ sqrt(eps()) * norm(D)
end

# Issue #139
@testset ExtendedTestSet "Matrix-vector products with SparseMatrix and SparseVector" begin
A = sprand(10, 10, 0.2)
b = sprand(10, 0.2)
opA = LinearOperator(A)
@test opA * b == A * b
@test transpose(opA) * b == transpose(A) * b
@test adjoint(opA) * b == adjoint(A) * b
end
end

test_linop()