Daniel Pinyol

In julia 11.0r4 ```julia using Test, LaTeXStrings detect_ambiguities(Base, Core, LaTeXStrings; recursive=true) ``` ``` Skipping Base.cwstring 1-element Vector{Tuple{Method, Method}}: (write(io::IO, s::LaTeXString) @ LaTeXStrings ~/.julia/packages/LaTeXStrings/ZtSdh/src/LaTeXStrings.jl:94, write(io::Base.AnnotatedIOBuffer, x::AbstractString) @ Base strings/annotated.jl:491) ```

Fix keytype and zero #574

7

Fixes https://github.com/JuliaSparse/SparseArrays.jl/issues/574

```julia julia> using SparseArrays julia> v=spzeros(Int,Int32,10) 10-element SparseVector{Int64, Int32} with 0 stored entries julia> keytype(v) Int64 julia> zero(v) 10-element SparseVector{Int64, Int64} with 0 stored entries ```

hash(SparseVector) very slow

1

```julia const sp=spzeros(Int,1_000_000_000) 1000000000-element SparseVector{Int64, Int64} with 0 stored entries @btime hash(sp) 182.608 ms (0 allocations: 0 bytes) ``` I can create a PR with the alternative implementation if you...

It looks like `isequal(SparseVector)` loops along all dimensions instead of only the non zero ones. Comparing 2 instances which just contain 1 value take 1.5s ```julia julia> using SparseArrays julia>...

I guess it should check also the inner modules, maybe with a "recursive" parameter

prepend! allocates when X is empty even when there's capacity again

2

https://github.com/JuliaLang/julia/issues/56029 fixed allocations when inserting at the beginning of a vector even when there was capacity. However, the fix only works with "large" vectors. With small ones (5 or less...

ExceptionInfo.value not available if pytest.raises fails

7

When pytest.raise fails, exc_info.value is not available. ``` def test(): try: with pytest.raises(AssertionError) as exc_info: raise Exception() except Exception as e: exc_info.value pass ``` I get this error ``` test_testchecks.py:33:...

When running mypy on a project using tensorboardX, you get this error ``` error: Skipping analyzing "tensorboardX": module is installed, but missing library stubs or py.typed marker [import-untyped] from tensorboardX...