Jishnu Bhattacharya

remove coverage files from Julia's share directory

1

Fixes #9, but I'm not sure if this is entirely desirable

Bump dependency compat

4

These versions should be compatible, as the breaking changes only impacts `ApproxFun.jl`

Multiplication of FFT plan and AbstractArray stack-overflows

1

https://github.com/JuliaMath/AbstractFFTs.jl/blob/03ef58b6f1fc033cba090a24c94837013b9f3ab0/src/definitions.jl#L220 This may be self-recursive MWE ```julia julia> f = ApproxFunBase.plan_transform(Chebyshev(), rand(4)) FastTransforms.ChebyshevTransformPlan{Float64, 1, (5,), false, 1, UnitRange{Int64}}(FFTW r2r REDFT10 plan for 4-element array of Float64 (redft10e-r2hc-4 (rdft-r2hc-direct-r2c-4 "r2cf_4"))) julia>...

Compact show for Frequencies

6

This PR specializes `show` for `Frequencies`. It's a workaround for https://github.com/JuliaLang/julia/issues/39963, and in general this makes the output more readable while displaying frequencies obtained from `fftfreq` and `rfftfreq`. On master...

This matches the docstring of `complex`, so now ```julia julia> typeof(complex(zero(Dual128))) == complex(Dual128) true ```

Fix indexing boolean ranges with empty ranges

4

Fixes #40760 Now ```julia julia> (false:false)[1:0] true:false julia> (false:true:false)[1:1:0] 0:1:-1 julia> (false:true:false)[1:0] 0:1:-1 ```

julia ```julia julia> @variables a,b,c,d,e,f,g,h,i,j 10-element Vector{Num}: a b c d e f g h i j julia> [a b; e h] 2×2 Matrix{Num}: a b e h julia> VERSION...

modules in test files

1

This avoids namespace issues, and makes each file independently executable.

Update checkout action to v3 in CI

1

Copied from https://github.com/actions/checkout, hopefully it'll get around the outdated node.js12 jobs warning

type assertion in convert

1

Since something similar is implemented on Julia nightly now, it seems to be the preferred way forward. After this, the following is inferred: ```julia julia> @inferred (x -> convert(OffsetArray{Float64,2,Matrix{Float64}}, x[1]))(Any[zeros(1,1)])...