get msprime working

[ajrominger]

• get conda env working and figure out a plan for msprime install #118
• figure out best way to deal with scale, generations, and J when updating phylo and handing off to msprime #153
• fix saturated pi issue
• make sure pi and Tajima D are assigned to right spp
• check that immigration is correct and not saturating pi
• check that harmonic mean is being calculate correct when there are 0s in abundance history
• issue with how to deal with new species added by speciation (see below)

[ajrominger]

this fails sometimes:

test_that("untbParams models run without error", {
    p <- untbParams(individuals_local = 100, 
                    individuals_meta = 1000, 
                    species_meta = 100, 
                    speciation = 0, 
                    dispersal_prob = 0.1, 
                    init_type = 'oceanic_island',
                    niter = 300000, niterTimestep = 10000)
    
    expect_no_error(runRole(roleModel(p)))
})

with output

The provenance information for the resulting tree sequence is 8.51MB. This is nothing to worry about as provenance is a good thing to have, but if you want to save this memory/storage space you can disable provenance recording by setting record_provenance=False
── Failure: untbParams models run without error ──────────────────────────────────────────────────────────────
Expected `runRole(roleModel(p))` to run without any errors.
ℹ Actually got a <msprime._msprime.LibraryError> with text:
  msprime._msprime.LibraryError: The derived population in a population split must be active
  Run `reticulate::py_last_error()` for details.

[ajrominger]

New species added by speciation

Might need to roll back changes to version prior to 6/18

Issue is that new species on a lineage deactivate the parent, so if t1 gives rise to t3 then there ancestor (which was also called t1) becomes inactive.

Could prune tree to just those taxa initially present and then graft on populations for the new species. The issue is that complex speciation patterns can be real tricky like a new species becomes the parent to another new species, or worse yet a new species becomes a parent to a new species and then goes extinct.

If can, the best approach might be to be able to do something like take samples on internal branches and track any corresponding name changes