improvements to A_mul_B! section of the manual #16702
Conversation
|
|
||
| Usually a function has 4 methods defined, one each for ``Float64``, | ||
| ``Float32``, ``Complex128`` and ``Complex64`` arrays. | ||
| In Julia and most other scientific computing, linear-algebra operations are |
There was a problem hiding this comment.
most other scientific computing ...
environments? languages?
There was a problem hiding this comment.
Actually, I meant "scientific computing" in general, though I should qualify that as dense linear algebra only.
There was a problem hiding this comment.
Unless you happen to use eigen?
There was a problem hiding this comment.
"Most," not all. Sure, there are Eigen, Blaze, Armadillo, and a few others, but these are by far in the minority as far as I can tell.
There was a problem hiding this comment.
I changed it to "In Julia (as in much of scientific computation), dense linear-algebra ..."
stevengj
force-pushed
the
A_mul_doc
branch
2 times, most recently
from
1fee9d0 to
751459d
Compare
kshyatt
added
docs
|
AppVeyor failure seems unrelated; some kind of timeout. |
|
@andreasnoack, do you have an opinion on this? I'm not sure what your |
|
After other array changes, I'll have to do a big rebase anyway so if you can rebase this I think we should merge soon. |
…t Ac_ldiv_B\! and At_ldiv_B!
|
Rebased. |
|
|
||
| .. Docstring generated from Julia source | ||
|
|
||
| Compute ``A B`` in-place and store the result in ``Y``\ , returning the result. If only two arguments are passed, then ``A_ldiv_B!(A, B)`` overwrites ``B`` with the result. |
There was a problem hiding this comment.
genstdlib ate the backslash here
Fix typo noted by @tkelman (#16702 (comment))
Fix typo noted by @tkelman (JuliaLang#16702 (comment))
A student of mine mentioned to me that he had trouble finding the documentation for
A_mul_B!in the manual because it was listed under "math" and not "linear algebra". In this PR, I fixed that and then made a few related changes:A_ldiv_B!was exported but not documented, so I added documentation and also added an optional 3-argument version to make it act a bit more likeA_mul_B!At_ldiv_Bwas exported, but notAt_ldiv_B!, as noted in rename KSPSolveTranspose! to Base.LinAlg.At_ldiv_B! JuliaParallel/PETSc.jl#70 (closes Export At_ldiv_B! #16425). Similarly forAc_ldiv_B!. I understand that @andreasnoack has some longer-term plans forldiv!, but in the meantime the current state of affairs seems pretty inconsistent.