You can find some help in the docstring, by typing ?fft
. And there is documentation with links in FFTW.jl: https://juliamath.github.io/FFTW.jl/stable/fft.html
Ok, docstring is good. But the page you linked doesnāt contain fft
documentation. It only refers to AbstractFFTs.jl.
Iām not saying that itās not traceable, Iām just saying that itās (too) difficult for new users for such a (arguably) basic functionality.
Huh? It absolutely has both a link to AbstractFFTs.jl and several screenfuls of documentation. Thereās something weird going on hereā¦
Correct: ā¦equations ecosystem (DifferntialEquations.jl),ā¦
By: ā¦equations ecosystem (DifferentialEquations.jl),ā¦
The situation is not perfect and there is an urgent need to have a centralized place where one can search for package functionality. I think everybody agrees on that and is encouraged to contribute to a better experience.
But: In any ways, the Julia documentation should not be the place to provide links to functionality that Julia does not provide. It should just document what is inside.
What we also need are metapackages that group smaller packages together. In case of the FFT this could be a SignalProcessing
package. The metapackage can reexport the fft function so that the user does not need to know the exact name of the package. Much like Toolboxes in Matlab.
Threreās already DSP.jl. It already depends on FFTW.jl, so re-exporting fft
and friends wouldnāt add a dependency.
Yes, DSP.jl could be the home for such a signal processing functionality. But currently fft is not reexported. The Images.jl packages are a good example where basics functionality is grouped into a larger package. Images.jl is mostly a metapackage and nobody has to remember that there is ImagesCore.jl and so on.
Great
Now we just need to wait for Julia Pro based in Julia 1.0, and all packages working properly.