Draft: Add rekey method

[glennmoy]

Probably should be overloading the methods introduced in mcabbott/AxisKeys.jl#63

I'll note a potentially concerning edge-case here, which existed prior to this but is maybe exposed more blatantly by it.
If the parent distribution is constructed with a KeyedArray but the KeyedDistribution is given a new set of keys there is no expectation that these should match, e.g.

ka = KeyedArray(ones(3); id=[:a, :b, :c])

kd = KeyedDistribution(MvNormal(ka), [:x, :y, :z])

[codecov]

Codecov Report

Merging #15 (0924fb5) into main (15926a2) will increase coverage by 0.03%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main      #15      +/-   ##
==========================================
+ Coverage   98.00%   98.03%   +0.03%     
==========================================
  Files           1        1              
  Lines          50       51       +1     
==========================================
+ Hits           49       50       +1     
  Misses          1        1              

Impacted Files	Coverage Δ
src/KeyedDistributions.jl	98.03% <100.00%> (+0.03%)	⬆️

---

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 15926a2...0924fb5. Read the comment docs.

[bencottier]

Looks good to me

[bencottier]

I'll note a potentially concerning edge-case here

Good catch, I'm writing an issue. It's undocumented so not a bug, but would be useful to check this condition.

#17