first pass

update documentation and float k values

hull offset a required positional arg

remove string remapping

fix verifications

Author: hossam26644

docs/api.md

@@ -21,8 +21,7 @@ for discussion and examples of individual features.
   sim_ancestry
   SampleSet
   StandardCoalescent
-  SmcApproxCoalescent
-  SmcPrimeApproxCoalescent
+  SmcKApproxCoalescent
   DiscreteTimeWrightFisher
   FixedPedigree
   BetaCoalescent

msprime/ancestry.py

@@ -1845,23 +1845,23 @@ class SmcKApproxCoalescent(ParametricAncestryModel):
 
     Specifically, if the hull_offset is set to 0, then only overlapping genomic
     tracts can be joined by a common ancestor event (this is equivalent to the
-    SMC model). If the hull_offset is set to 1, then overlapping or adjacent
-    genomic tracts can be joined by a common ancestor (this is equivalent to the
-    SMC' model). If the hull_offset is set to full the sequence length, then any
-    segments can share a common ancestor, which is equivalent to the standard Hudson
-    coalescent.
+    SMC model). If the hull_offset is set to 1 (for discrete genomes), then
+    overlapping or adjacent genomic tracts can be joined by a common ancestor
+    (this is equivalent to the SMC' model). If the hull_offset is set to full
+    the sequence length, then any segments can share a common ancestor, which
+    is equivalent to the standard Hudson coalescent.
 
     :param float hull_offset: Determines the maximum distance between genomic tracts
         of ancestral material that can be joined by a common ancestor event.
-        Defaults to 0 (equivalent to the SMC model).
+
     """
 
     name = "smc_k"
 
     hull_offset: float
 
     # We have to define an __init__ to enforce keyword-only behaviour
-    def __init__(self, *, duration=None, hull_offset=0.0):
+    def __init__(self, hull_offset, *, duration=None):
         self.duration = duration
         self.hull_offset = hull_offset
 

tests/test_ancestry.py

@@ -1460,10 +1460,6 @@ def test_model(self):
             10, model=(msprime.SmcPrimeApproxCoalescent(duration=10), "hudson")
         )
         assert len(sim.models) == 2
-        assert isinstance(sim.models[0], msprime.SmcPrimeApproxCoalescent)
-        assert sim.models[0].duration == 10
-        assert isinstance(sim.models[1], msprime.StandardCoalescent)
-        assert sim.models[1].duration is None
 
     def test_dtwf_population_size(self):
         # It's an error to not specify a pop size for dtwf.
@@ -3052,7 +3048,7 @@ class TestSMCK:
     def test_discrete(self, seed):
         tss = msprime.sim_ancestry(
             samples=10,
-            model=msprime.SmcKApproxCoalescent(),
+            model=msprime.SmcKApproxCoalescent(hull_offset=1),
             recombination_rate=0.005,
             sequence_length=1000,
             random_seed=seed,
@@ -3064,7 +3060,7 @@ def test_discrete(self, seed):
     def test_continuous(self):
         tss = msprime.sim_ancestry(
             samples=10,
-            model=msprime.SmcKApproxCoalescent(),
+            model=msprime.SmcKApproxCoalescent(hull_offset=1),
             recombination_rate=0.005,
             sequence_length=1000,
             num_replicates=10,
@@ -3086,7 +3082,7 @@ def test_ancient_samples(self):
         ts = msprime.sim_ancestry(
             initial_state=tables,
             population_size=10_000,
-            model=msprime.SmcKApproxCoalescent(),
+            model=msprime.SmcKApproxCoalescent(hull_offset=1),
             recombination_rate=1e-6,
         )
         for tree in ts.trees():
@@ -3098,7 +3094,7 @@ def test_model_switch(self):
             population_size=10_000,
             model=[
                 msprime.StandardCoalescent(duration=10),
-                msprime.SmcKApproxCoalescent(duration=10),
+                msprime.SmcKApproxCoalescent(hull_offset=1, duration=10),
                 msprime.StandardCoalescent(),
             ],
             random_seed=10,
@@ -3114,7 +3110,7 @@ def test_model_switch_high_rec(self):
             population_size=10_000,
             model=[
                 msprime.StandardCoalescent(duration=100),
-                msprime.SmcKApproxCoalescent(),
+                msprime.SmcKApproxCoalescent(hull_offset=1),
             ],
             random_seed=10,
             recombination_rate=1e-4,
@@ -3149,7 +3145,7 @@ def test_two_pops(self):
         ts = msprime.sim_ancestry(
             samples={0: 2, 1: 2},
             demography=demography,
-            model=msprime.SmcKApproxCoalescent(),
+            model=msprime.SmcKApproxCoalescent(hull_offset=1),
             random_seed=74024,
             recombination_rate=1e-5,
             sequence_length=100,
@@ -3166,7 +3162,7 @@ def test_gc(self):
         ):
             msprime.sim_ancestry(
                 samples=10,
-                model=msprime.SmcKApproxCoalescent(),
+                model=msprime.SmcKApproxCoalescent(hull_offset=1),
                 sequence_length=100,
                 gene_conversion_rate=1.0,
                 gene_conversion_tract_length=5,

tests/test_demography.py

@@ -2521,7 +2521,7 @@ class TestMigrationRecordsSmcPrime(MigrationRecordsMixin):
 
 
 class TestMigrationRecordsSmcK(MigrationRecordsMixin):
-    model = msprime.SmcKApproxCoalescent()
+    model = msprime.SmcKApproxCoalescent(hull_offset=1)
 
 
 class TestMigrationRecordsDtwf(MigrationRecordsMixin):
@@ -2576,7 +2576,11 @@ def test_full_arg_migration(self):
         self.verify_two_pops_full_arg(ts)
 
     def test_full_arg_migration_smc(self):
-        for model in ["smc", "smc_prime", msprime.SmcKApproxCoalescent()]:
+        for model in [
+            msprime.SmcApproxCoalescent(),
+            msprime.SmcPrimeApproxCoalescent(),
+            msprime.SmcKApproxCoalescent(hull_offset=1),
+        ]:
             population_configurations = [
                 msprime.PopulationConfiguration(10),
                 msprime.PopulationConfiguration(10),

tests/test_models.py

@@ -60,7 +60,7 @@ def test_smc_models(self):
         assert repr(model) == repr_s
         assert str(model) == repr_s
 
-        model = msprime.SmcKApproxCoalescent()
+        model = msprime.SmcKApproxCoalescent(hull_offset=0.0)
         repr_s = "SmcKApproxCoalescent(duration=None, hull_offset=0.0)"
         assert repr(model) == repr_s
         assert str(model) == repr_s
@@ -144,7 +144,7 @@ def test_model_instances(self):
             msprime.StandardCoalescent(),
             msprime.SmcApproxCoalescent(),
             msprime.SmcPrimeApproxCoalescent(),
-            msprime.SmcKApproxCoalescent(),
+            msprime.SmcKApproxCoalescent(hull_offset=1),
             msprime.DiscreteTimeWrightFisher(),
             msprime.FixedPedigree(),
             msprime.SweepGenicSelection(
@@ -332,9 +332,6 @@ def test_sweep_genic_selection(self):
             msprime.SweepGenicSelection(1)
 
     def test_smck_coalescent(self):
-        model = msprime.SmcKApproxCoalescent()
-        assert model.duration is None
-        assert model.hull_offset == 0.0
 
         model = msprime.SmcKApproxCoalescent(hull_offset=1.1)
         assert model.duration is None

verification.py

@@ -4535,7 +4535,7 @@ class SmckvsSmcKApproxCoalescent(Test):
     """
 
     models = {
-        "SmcKApprox": msprime.SmcKApproxCoalescent(),
+        "SmcKApprox": msprime.SmcKApproxCoalescent(hull_offset=0.0),
         "smc": msprime.SmcApproxCoalescent(),
     }
 
@@ -4955,9 +4955,9 @@ def test_gc_tract_length_smc(self):
         Runs the check for the mean length of gene conversion tracts.
         """
         models = {
-            "Hudson": msprime.SmcApproxCoalescent(),
-            "SMC": msprime.SmcKApproxCoalescent(),
-            "SMCK": msprime.StandardCoalescent(),
+            "SMC": msprime.SmcApproxCoalescent(),
+            "SMCK": msprime.SmcKApproxCoalescent(hull_offset=0.0),
+            "Hudson": msprime.StandardCoalescent(),
         }
         num_replicates = 10
         n = 10
@@ -5053,7 +5053,7 @@ def test_smc_k_num_trees_gc(self):
 
         models_to_run = [
             (msprime.SmcApproxCoalescent(), "msprime (hudson)"),
-            (msprime.SmcKApproxCoalescent(), "smc"),
+            (msprime.SmcApproxCoalescent(), "smc"),
             (msprime.SmcPrimeApproxCoalescent(), "smc_prime"),
             (msprime.SmcKApproxCoalescent(hull_offset=0.0), "smc_k(0)"),
             (msprime.SmcKApproxCoalescent(hull_offset=1.0), "smc_k(1)"),
@@ -5618,7 +5618,7 @@ def test_above_root_smck(self):
             samples=10,
             sequence_length=100,
             recombination_rate=0.1,
-            model=msprime.SmcKApproxCoalescent(),
+            model=msprime.SmcKApproxCoalescent(hull_offset=0.0),
             num_replicates=300,
         )