do not simulate null ancestry; closes #357

[petrelharp]

The downside here is that the resulting tree sequence can't be loaded back into SLiM. We could go and put the 'sample' flags back after recapitation, but it just seems weird to have them there anyhow. All this will be changing with the new model.

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 21.42857% with 11 lines in your changes missing coverage. Please review.

Project coverage is 94.01%. Comparing base (46809d4) to head (658f9fd).

Files with missing lines	Patch %	Lines
pyslim/methods.py	21.42%	9 Missing and 2 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #359      +/-   ##
==========================================
- Coverage   95.65%   94.01%   -1.64%     
==========================================
  Files           8        8              
  Lines         621      635      +14     
  Branches      131      136       +5     
==========================================
+ Hits          594      597       +3     
- Misses         18       27       +9     
- Partials        9       11       +2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.