How to pass variables to several levels up without going through headings (arguments) in functions

[fmontans]

@goerz

I would even say that Julia and modern Fortran are so similar that I’ve been able to port large swaths of Fortran code by copying it to a .jl file and doing some very basic search-and-replace to adjust the syntax.

The same thinking as fortran is probably why I like so much Julia. Aside greek symbols, subindices and defining operator symbols. Extremely easy to read, hence to understand and maintain. And as a plus, getting rid of Matlab. All this is why I want to move my code to Julia (apart from engaging my much-younger students into it).

Again, thank you all a lot for all the support to the novice!

[nilshg]

I’m not in the field and to be honest don’t even know what is is (fine elements? Aren’t they all? I’ve never met an infinite one!) but I thought I’d mention that there’s a well maintained library/ecosystem here

GitHub - PetrKryslUCSD/FinEtools.jl: Finite Element tools in Julia

which might be of interest. (I’d tag Petr as well if the Discourse tagging thing would work on mobile!)

[bertschi]

Reminds me of this Ed Post quote from '82:

The determined Real Programmer can write FORTRAN programs in any language.

On the other hand, I probably just tend to write Common Lisp or Haskell in Julia

const diam = splat((:))∘extrema

[Eben60]

@bertschi

Collect related parameters into a struct and just pass it as a single argument.

If the only purpose is just to pass a large number of arguments, I pack them into a NamedTuple. Which can be extended (by merging), mutated (by merging), unpacked by name or by position, or passed as kwargs.

[stevengj]

@nilshg

but I thought I’d mention that there’s a well maintained library/ecosystem here: FinEtools.jl

(There is also Gridap.jl and some other packages … I seem to remember a summary of Julia FEM packages somewhere, but I can’t find it at the moment.)