[WIP] Use static arrays

[giordano]

Fix #45.

This is marked as work-in-progress because I'm not sure I made all possible changes. Feel free to report any missing array to change.

[codecov-io]

Codecov Report

Merging #46 into master will increase coverage by 0.06%. The diff coverage is 100%.

@@            Coverage Diff             @@
##           master      #46      +/-   ##
==========================================
+ Coverage   99.81%   99.87%   +0.06%     
==========================================
  Files          76       76              
  Lines        1581     1554      -27     
==========================================
- Hits         1578     1552      -26     
+ Misses          3        2       -1

Impacted Files	Coverage Δ
src/common.jl	100% <ø> (ø)	:arrow_up:
src/gal_uvw.jl	100% <ø> (ø)	:arrow_up:
src/precess_cd.jl	100% <100%> (ø)	:arrow_up:
src/helio.jl	100% <100%> (ø)	:arrow_up:
src/moonpos.jl	100% <100%> (ø)	:arrow_up:
src/nutate.jl	100% <100%> (ø)	:arrow_up:
src/premat.jl	100% <100%> (ø)	:arrow_up:
src/mag2geo.jl	100% <100%> (ø)	:arrow_up:
src/baryvel.jl	100% <100%> (ø)	:arrow_up:
src/sixty.jl	100% <100%> (ø)	:arrow_up:
... and 22 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update c8358fc...170d785. Read the comment docs.

[TestSubjector]

Perhaps the following changes?

In bprecess.jl, change line 42 to

s = Vector{T}(s1)

Additionally, since premat() returns an array; either convert it to static or convert the variable that receives the return value of premat (in precess.jl and baryvel.jl) to static?

Small question :- Does StaticArrays give a significant boost to the performance of the code?

[giordano]

Thanks for looking into it!

In bprecess.jl, change line 42 to

s = Vector{T}(s1)

s will be changed, doing what you suggest would change s1 as well, this is what I wanted to avoid with the copy. See this example:

julia> s = randn(3)
3-element Array{Float64,1}:
  0.593006
 -1.79782 
  0.592398

julia> s1 = Vector{Float64}(s)
3-element Array{Float64,1}:
  0.593006
 -1.79782 
  0.592398

julia> s1[1] = -5
-5

julia> s
3-element Array{Float64,1}:
 -5.0     
 -1.79782 
  0.592398

Additionally, since premat() returns an array; either convert it to static or convert the variable that receives the return value of premat (in precess.jl and baryvel.jl) to static?

Not sure what you refer to, the PR already makes the array returned by premat static: https://github.com/JuliaAstro/AstroLib.jl/pull/46/files#diff-3466cd565556cdaaf4ea10f869aa4961

[TestSubjector]

Not sure what you refer to, the PR already makes the array returned by premat static: https://github.com/JuliaAstro/AstroLib.jl/pull/46/files#diff-3466cd565556cdaaf4ea10f869aa4961

Ah, didn't notice that. I was referring to the same.

I think almost everything is covered by the PR then.

[giordano]

I think almost everything is covered by the PR then.

I don't expect to have covered everything, but almost everything is a good result :-) Thanks!

[giordano]

Holding back the PR because of https://github.com/JuliaArrays/StaticArrays.jl/issues/321 which makes nutate type-unstable on Julia master https://github.com/JuliaArrays/StaticArrays.jl/issues/322 https://github.com/JuliaLang/julia/issues/24286

Edit: all issues resolved.