David Weber

Complex{Float32} is not a subtype of Float32, breaking conv of complex valued input/weights

3

Just a simple convolution with two complex matrices breaks: `conv(randn(Complex{Float32}, 25,25,1,3), randn(Complex{Float32}, 5,5,1,3))` The stated reason is that this line https://github.com/FluxML/NNlib.jl/blob/0d16973bab1260de045c1599ec3d12e5adac1d70/src/nnpack/interface.jl#L15 is trying to coerce a complex into a real...

maxpool NNPACK breaks nonzero padding with stride 1

3

As the title says, using `maxpool` on the cpu with nonzero padding and a stride of 1 results in an error. On a previous version, `maxpool` with these parameters worked...

Fixing exists by adding keywords

1

Exists doesn't work when you have an environment that's a string, such as if you have a different root environment. The first commit just addresses exists, while the second one...

Saving figure using export_fig using gcf

19

I'm attempting to use some old plotting code, and I can't seem to get export_fig to save. Here's a minimal example showing the behavior. ```julia using MATLAB y = linspace(2.,...

Source code docs links go to base Julia source code, not LineSearches.jl

1

Specifically, all the links on [this page](https://julianlsolvers.github.io/LineSearches.jl/stable/reference/linesearch.html) link to https://github.com/JuliaLang/julia/blob/0d713926f85dfa3e4e0962215b909b8e47e94f48/base/#L0-L12 with the last bit changing occasionally. I clicked on Edit on GitHub, but it appears the docs were created with...

Citations for the different methods

3

I have found this repository very useful, and would like to cite it and the methods it uses. However, as someone not directly involved in the evolutionary computing space, it...

Test errors: FloatingPoint not defined (julia 0.5.0)

2

I'm planning on using Regression to do multinomial least squares, but when I tried to load the command after `using Regression`, it wasn't found. Naturally, I tried to test the...

@allowscalar triggers try/catch error

6

I'm not sure if I should be filing this here or in CUDA, but doing temporary `@allowscalar`s lead to errors when taking a gradient. MWE: ``` using Zygote, CUDA CUDA.allowscalar(false)...

What happened: ``` > here::here("a", "", "b") [1] "~/a//b" ``` What I expected ``` > here::here("a", "", "b") [1] "~/a/b" ``` This is also a "feature" of `file.path`. As written...

Styler and lintr disagree on what happens after a linebreak

2

Styler auto-formats to: ```r if (TRUE || FALSE) { TRUE } ``` While linter prefers: ```r if (TRUE || FALSE) { TRUE } ``` I don't really have strong opinions...