parametrize SymTridiagonal on the wrapped vector type #23035
Conversation
fredrikekre
added
breaking
NEWS.md
Outdated
| @@ -73,8 +73,10 @@ This section lists changes that do not have deprecation warnings. | |||
| longer present. Use `first(R)` and `last(R)` to obtain | |||
| start/stop. ([#20974]) | |||
|
|
|||
| * The `Diagonal` type definition has changed from `Diagonal{T}` to | |||
| `Diagonal{T,V<:AbstractVector{T}}` ([#22718]). | |||
| * The `Diagonal`, `Bidiagonal` and `SymTridiagonal` type definitions have changed from | |||
There was a problem hiding this comment.
assuming that merges first? I'm still not so sure about that densifying constructor over there
base/linalg/tridiag.jl
Outdated
|
|
||
| function SymTridiagonal(dv::AbstractVector{Td}, ev::AbstractVector{Te}) where {Td,Te} | ||
| T = promote_type(Td,Te) | ||
| function SymTridiagonal(dv::AbstractVector{T}, ev::AbstractVector{T}) where T | ||
| SymTridiagonal(convert(Vector{T}, dv), convert(Vector{T}, ev)) |
There was a problem hiding this comment.
here as well, it seems you wouldn't want to do this any more
There was a problem hiding this comment.
Could you suggest an alternative promotion pattern to obtain the same type of both vectors? Perhaps over here: #22925 (comment)
fredrikekre
force-pushed
the
fe/symtridiag
branch
from
ef04e4b
to
f0c2bbe
Compare
base/deprecated.jl
Outdated
| @@ -1599,6 +1599,15 @@ end | |||
| # issue #6466 | |||
| # `write` on non-isbits arrays is deprecated in io.jl. | |||
|
|
|||
| # PR #23035 | |||
| # also uncomment constructor tests in test/linalg/tridiag.jl | |||
| function SymTridiagonal(dv::AbstractVector{T}, ev::AbstractVector{S}) where {T,S} | |||
There was a problem hiding this comment.
could also use a deprecation bullet in news (or could add it for both in whichever merges second)
fredrikekre
force-pushed
the
fe/symtridiag
branch
2 times, most recently
from
48732d8
to
a3d50cf
Compare
|
@nanosoldier |
fredrikekre
force-pushed
the
fe/symtridiag
branch
from
a3d50cf
to
538904c
Compare
|
Your benchmark job has completed - possible performance regressions were detected. A full report can be found here. cc @ararslan |
|
Let's try again |
|
Your benchmark job has completed - no performance regressions were detected. A full report can be found here. cc @ararslan |
SymTridiagonalversion of #22718 and #22925.