Fix imrotate for Irrational θ

[timholy]

I was just about to release 0.9.1 (in the wake of #148) and happened to try the test cases in #79, and got this:

julia> mod2pi(π)
ERROR: MethodError: no method matching rem2pi(::Irrational{:π}, ::RoundingMode{:Down})
Closest candidates are:
  rem2pi(::Float64, ::RoundingMode{:Down}) at math.jl:1039
  rem2pi(::Float32, ::RoundingMode) at math.jl:1098
  rem2pi(::Float16, ::RoundingMode) at math.jl:1099
  ...
Stacktrace:
 [1] mod2pi(x::Irrational{:π})
   @ Base.Math ./math.jl:1128

That would not have been a nice thing to release. I'm glad luck worked on my side this time 😅

[codecov]

Codecov Report

Merging #149 (0e3e44d) into master (516dbf6) will increase coverage by 0.23%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #149      +/-   ##
==========================================
+ Coverage   89.90%   90.14%   +0.23%     
==========================================
  Files           8        8              
  Lines         208      213       +5     
==========================================
+ Hits          187      192       +5     
  Misses         21       21              

Impacted Files	Coverage Δ
src/warp.jl	100.00% <100.00%> (ø)

---

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 516dbf6...0e3e44d. Read the comment docs.

[timholy]

I guess one other thing to think about: is the change in size of the returned images breaking? Should this be 0.10 instead?

[timholy]

Might be best to find out whether JuliaGeometry/Rotations.jl#163 meets with approval, and if so just require it.

[johnnychen94]

Unsure how it implies, but it looks good to me.

[timholy]

Unsure how it implies

Sorry, I'm not sure I know what you mean. Do you mean, you're not sure how that Rotations PR applies? We wouldn't need rotmtrx2 anymore. But given that master is a bit broken, perhaps it's best to go with this for now and rework this as needed later.

[timholy]

Any thoughts on the 0.9.1 vs 0.10 issue? That was the main thing I pinged you for (sorry, I know that wasn't clear).

[johnnychen94]

Do you mean, you're not sure how that Rotations PR applies?

Sorry for the inaccuracy, I meant I wasn't very sure how this PR co-works with JuliaLang/julia#42595 and JuliaGeometry/Rotations.jl#163. IIUC JuliaLang/julia#42595 doesn't improve this PR, am I right?

Any thoughts on the 0.9.1 vs 0.10 issue?

Ah, I missed the first comment. I think we can make it 0.9.1 here as it's a clear improvement from my point of view, can you also update the CHANGELOG.md? Thanks again for taking care of this!

[timholy]

Ah, gotcha. Yes, we might not even need to change Rotations if the Base PR goes through. I could put definitions into Compat and then they would be available for previous Julia versions.