MatFi

> ComponentArrays can have a heterogeneous mix of types now, so mixed units can work. Sadly `one(::ComponentVector)` is not implemented so it fails with implicit solvers

> it would be good if this was just a batch interface I definitely see the advantages of doing this, but keep in mind that some functions can vary a...

Profiling revealed that it is because `c` is not sorted. ```julia a = 0:.01:1 b = a .^ 2 c = sort(rand(10^6)) itp1 = LinearInterpolation((a,),b) # Regular Grid itp2 =...

In 1.7.2 i get: ```julia julia> promote_type(Float32, Gray{Float32}) Any ``` which is also bad. Looks like that supporting type instability ( #370 #424 ) is in general a bad idea...

You can try a `MonotonicInterpolation` like [SteffenMonotonicInterpolation](https://juliamath.github.io/Interpolations.jl/latest/api/#Interpolations.SteffenMonotonicInterpolation) , they are mostly of order 3 I think. Edit: they are only 1D.. sry for the noise.

Seems that the gradient is implemented not consistent. In case of a higher-order extrapolation, simply the gradient nearest to the edge of the area to be extrapolated is returned. Probably...

> As of now, filter objects and channels are reference counted and automatically deleted when the last consumer (measurement display, sink / export filter, waveform view, or filter/channel input) is...