Simeon David Schaub
Simeon David Schaub
``` julia> using StatsBase julia> using Zygote [ Info: Precompiling Zygote [e88e6eb3-aa80-5325-afca-941959d7151f] julia> Zygote.pullback(mean, 1:4, Weights(1:4)) ERROR: MethodError: objects of type UnitRange{Int64} are not callable Use square brackets [] for...
The gradient here should be one, not `nothing`: ```julia julia> using Zygote julia> Zygote.gradient(x -> (x:3)[1], 1.2) (nothing,) julia> 1.2:3 1.2:1.0:2.2 ``` It looks like the adjoint for `:` does...
Depends on https://github.com/JuliaLang/julia/pull/37902 and still needs to be made backwards-compatible. fixes #552 (at least on future Julia versions)
Steps to reproduce: Apply this patch to `base/range.jl` on the latest Julia master: ```diff diff --git a/base/range.jl b/base/range.jl index 5f30a31a0b..e004161551 100644 --- a/base/range.jl +++ b/base/range.jl @@ -318,7 +318,7 @@ be...
This is especially relevant for cross-compilation, since dwl uses `clock_gettime`.
MWE: ```julia julia> using Tullio, CUDA, KernelAbstractions, Zygote, FiniteDifferences julia> g(μ) = ifelse(μ == 4, 1, -1) g (generic function with 1 method) julia> push!(Tullio.BASE_NOGRAD, :g); julia> f2(w, k) =...
My main motivation behind this is that [MonteCarloMeasurements.jl](https://github.com/baggepinnen/MonteCarloMeasurements.jl) already overloads these, so I believe this would make for better interchangeability between both approaches. Personally, I also find them more convenient...
This corrects the printing of measurements in expressions, like `:(2 * (1 ± .2))`.
I have the following Python code I want to run within a LaTeX document: ```python @ASCII(expression='g / D * m') @LaTeX(name='\mathrm s', x_name=r'\mathrm m', parameter_names=(r'\mathrm g', r'\mathrm D'), expression=r'\frac{\mathrm g}{\mathrm...
I thought that file might be a good stress test of this package. It seems to do quite well. The first case is probably up to interpretation, since although the...