felixhorger

Hi All, Note that there another possibility of using complex numbers with LoopVectorization: Instead of having separate arrays containing real & imaginary parts, you can just reinterpret the complex array...

Hi, I am afraid this is more than stated above, e.g. saving a 8Gb Float64 array `jldsave("test.jld2", true; a=zeros(Float64, 1000, 1000000))` fails with the same error on my machine with...

You are right! The error goes all the way back to `CodecZlib.jl`. It can only take a ~4Gb block at a time. So either the codec is switched for large...

I think I solved the issue by modifying `CodecZlib.jl`, see this [pull request](https://github.com/JuliaIO/CodecZlib.jl/pull/62). I can now `jldsave` and `load` arrays larger than `typemax(UInt32)` bytes.

Hi Tim, sorry for taking so long to get back to you! I wrote a function which is faster than minimum_finite/maximum_finite from Images.jl by a factor of nearly 5. On...

Done! :) What do you think about using threads to speed up the value span determination?

Apologies for the late reply, a conference deadline kept me from pursuing this! Yes that would help! Basically I have a set of operators A, B, ... which are multiplied...

Thanks for working towards a solution! I tried it on v3.10.0-DEV but it still allocates, here is a MWE to reproduce ``` using LinearMaps N = 100_000_000 function plan() y...

Yes, that works like a charm, thanks! :) When I test it with the above snippet, the `@assert` statements are both true and I can only observe a small 80...

Interesting, as far as I can see it does use the same commands, just run from outside WSL, it's good to have that solution available as well, plus it might...