Check for valid input in ARG likelihood

[JereKoskela]

This outputs a readable error message when ARGs are requested for trees which aren't binary.

I haven't put in a check for unary nodes because those will legitimately appear on trees due to RE nodes.

Closes #2107

[codecov]

Codecov Report

Merging #2114 (cc6025a) into main (360dfec) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main    #2114   +/-   ##
=======================================
  Coverage   91.29%   91.29%           
=======================================
  Files          19       19           
  Lines       10852    10856    +4     
  Branches     2222     2224    +2     
=======================================
+ Hits         9907     9911    +4     
  Misses        516      516           
  Partials      429      429           

Flag	Coverage Δ
C	91.29% <100.00%> (+<0.01%)	⬆️
python	98.73% <100.00%> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
msprime/likelihood.py	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 360dfec...cc6025a. Read the comment docs.

[jeromekelleher]

Can you rebase once #2116 is merged please @JereKoskela?

[JereKoskela]

Done @jeromekelleher

[jeromekelleher]

Needs a test or two, can be something quite simple. Probably easiest to hand construct a few tree sequences that we'd like to reject.

[JereKoskela]

I've added a test to cover the ValueError. It's only one, but any further tests seemed redundant since they'd just be more elaborate (and less human-readable) versions of the same scenario. Let me know if there's something else in particular we should be covering @jeromekelleher.

[jeromekelleher]

LGTM