def test_zone_taxon_range_mask(zone_example_grid):
mg, z = zone_example_grid
# Create a zone for time 0.
ids = [9, 10, 11, 12]
z[ids] = 1
se = SpeciesEvolver(mg)
sc = ZoneController(mg, zone_func)
taxa = sc.populate_zones_uniformly(1)
se.track_taxa(taxa)
expected_mask = np.zeros(mg.number_of_nodes, bool)
expected_mask[ids] = True
np.testing.assert_array_equal(taxa[0].range_mask, expected_mask)
# Remove extent so taxa range mask is all False.
z[ids] = 0
sc.run_one_step(1)
se.run_one_step(1)
expected_mask = np.zeros(mg.number_of_nodes, bool)
np.testing.assert_array_equal(taxa[0].range_mask, expected_mask)
def test_time_to_allopatric_speciation(zone_example_grid):
mg, z = zone_example_grid
# Create a zone for time 0.
z[[9, 10, 11, 12]] = 1
se = SpeciesEvolver(mg)
sc = ZoneController(mg, zone_func)
taxa = sc.populate_zones_uniformly(1, time_to_allopatric_speciation=20)
se.track_taxa(taxa)
z[[11]] = 0
while len(se.taxa_data_frame) == 1:
sc.run_one_step(10)
se.run_one_step(10)
expected_df = pd.DataFrame(
{
"pid": [np.nan, 0],
"type": 2 * [ZoneTaxon.__name__],
"t_first": [0, 30],
"t_final": 2 * [np.nan],
},
index=[0, 1],
)
expected_df.index.name = "tid"
expected_df["pid"] = expected_df["pid"].astype("Int64")
expected_df["t_final"] = expected_df["t_final"].astype("Int64")
pd.testing.assert_frame_equal(se.taxa_data_frame, expected_df, check_like=True)
def test_allopatric_wait_time(zone_example_grid):
mg, z = zone_example_grid
# Create a zone for time 0.
z[[9, 10, 11, 12]] = 1
se = SpeciesEvolver(mg)
sc = ZoneController(mg, zone_func)
taxa = sc.populate_zones_uniformly(1, allopatric_wait_time=20)
se.track_taxa(taxa)
z[[11]] = 0
while len(se.taxa_data_frame) == 1:
sc.run_one_step(10)
se.run_one_step(10)
expected_df = pd.DataFrame(
{
'appeared': [0, 30, 30],
'latest_time': [30, 30, 30],
'extant': [False, True, True]
},
index=[0, 1, 2])
pd.testing.assert_frame_equal(se.taxa_data_frame,
expected_df,
check_like=True)
def test_none_to_none(zone_example_grid):
mg, z = zone_example_grid
sc = ZoneController(mg, zone_func)
sc.run_one_step(1)
np.testing.assert_array_equal(len(sc.zones), 0)
def test_zone_func_kwargs(zone_example_grid):
mg, z = zone_example_grid
z[mg.core_nodes] = 1
sc = ZoneController(mg, zone_func_with_vars, var1=1, var2=2)
expected_mask = np.array(
[
False,
False,
False,
False,
False,
False,
False,
False,
False,
False,
False,
False,
False,
False,
False,
True,
True,
True,
True,
True,
False,
False,
True,
True,
True,
True,
True,
False,
False,
False,
False,
False,
False,
False,
False,
]
)
np.testing.assert_array_equal(sc.zones[0].mask, expected_mask)
z[-2:] = 0
expected_mask[-2:] = False
sc.run_one_step(10)
np.testing.assert_array_equal(sc.zones[0].mask, expected_mask)
def test_many_to_many(zone_example_grid):
mg, z = zone_example_grid
# Create two zones for time 0.
z[[10, 12, 17, 19, 24, 26]] = 1
se = SpeciesEvolver(mg)
sc = ZoneController(mg, zone_func)
taxa = sc.populate_zones_uniformly(1)
se.track_taxa(taxa)
expected_df = pd.DataFrame({
'time': [0],
'zones': [2],
'fragmentations': [np.nan],
'captures': [np.nan],
'area_captured_sum': [np.nan],
'area_captured_max': [np.nan]
})
pd.testing.assert_frame_equal(sc.record_data_frame,
expected_df,
check_like=True)
np.testing.assert_equal(len(se.get_taxon_objects(time=0)), 2)
# Modify elevation such that two zones each overlap the original two zones.
z[[17, 19]] = 0
z[[11, 25]] = 1
sc.run_one_step(1)
se.run_one_step(1)
np.testing.assert_equal(len(sc.zones), 2)
for z in sc.zones:
np.testing.assert_equal(z._conn_type, zn.Connection.MANY_TO_MANY)
expected_df = pd.DataFrame({
'time': [0, 1],
'zones': [2, 2],
'fragmentations': [np.nan, 0],
'captures': [np.nan, 2],
'area_captured_sum': [np.nan, 24],
'area_captured_max': [np.nan, 12]
})
pd.testing.assert_frame_equal(sc.record_data_frame,
expected_df,
check_like=True)
np.testing.assert_equal(
len(se.get_taxon_objects(extant_at_latest_time=True)), 4)
def test_one_to_many(zone_example_grid):
mg, z = zone_example_grid
# Create a zone for time 0.
z[[9, 10, 11, 12]] = 1
se = SpeciesEvolver(mg)
sc = ZoneController(mg, zone_func)
taxa = sc.populate_zones_uniformly(1)
se.track_taxa(taxa)
expected_df = pd.DataFrame(
{
"time": [0],
"zones": [1],
"fragmentations": [np.nan],
"captures": [np.nan],
"area_captured_sum": [np.nan],
"area_captured_max": [np.nan],
}
)
pd.testing.assert_frame_equal(sc.record_data_frame, expected_df, check_like=True)
np.testing.assert_equal(len(se.get_extant_taxon_objects(time=0)), 1)
# Break the zone in two for time 1.
z[11] = 0
sc.run_one_step(1)
se.run_one_step(1)
np.testing.assert_equal(len(sc.zones), 2)
np.testing.assert_equal(
set([z._conn_type for z in sc.zones]), set([None, zn.Connection.ONE_TO_MANY])
)
expected_df = pd.DataFrame(
{
"time": [0, 1],
"zones": [1, 2],
"fragmentations": [np.nan, 2],
"captures": [np.nan, 0],
"area_captured_sum": [np.nan, 0],
"area_captured_max": [np.nan, 0],
}
)
pd.testing.assert_frame_equal(sc.record_data_frame, expected_df, check_like=True)
np.testing.assert_equal(len(se.get_extant_taxon_objects()), 2)
def test_one_to_one(zone_example_grid):
mg, z = zone_example_grid
# Create a zone for time 0.
z[[9, 10, 11, 12]] = 1
se = SpeciesEvolver(mg)
sc = ZoneController(mg, zone_func)
taxa = sc.populate_zones_uniformly(1)
se.track_taxa(taxa)
np.testing.assert_equal(len(sc.zones), 1)
zone = sc.zones[0]
expected_df = pd.DataFrame(
{
"time": [0],
"zones": [1],
"fragmentations": [np.nan],
"captures": [np.nan],
"area_captured_sum": [np.nan],
"area_captured_max": [np.nan],
}
)
pd.testing.assert_frame_equal(sc.record_data_frame, expected_df, check_like=True)
# Modify elevation, although there is still one zone in time 1.
z[[11, 12]] = 0
sc.run_one_step(1)
se.run_one_step(1)
np.testing.assert_equal(len(sc.zones), 1)
np.testing.assert_equal(zone, sc.zones[0])
np.testing.assert_equal(sc.zones[0]._conn_type, zn.Connection.ONE_TO_ONE)
expected_df = pd.DataFrame(
{
"time": [0, 1],
"zones": [1, 1],
"fragmentations": [np.nan, 0],
"captures": [np.nan, 0],
"area_captured_sum": [np.nan, 0],
"area_captured_max": [np.nan, 0],
}
)
pd.testing.assert_frame_equal(sc.record_data_frame, expected_df, check_like=True)
np.testing.assert_equal(len(se.get_extant_taxon_objects(time=1)), 1)
def test_one_to_none(zone_example_grid):
mg, z = zone_example_grid
# Create a zone for time 0.
z[[9, 10, 11, 12]] = 1
se = SpeciesEvolver(mg)
sc = ZoneController(mg, zone_func)
taxa = sc.populate_zones_uniformly(1)
se.track_taxa(taxa)
np.testing.assert_equal(len(sc.zones), 1)
zone = sc.zones[0]
expected_df = pd.DataFrame(
{
"time": [0],
"zones": [1],
"fragmentations": [np.nan],
"captures": [np.nan],
"area_captured_sum": [np.nan],
"area_captured_max": [np.nan],
}
)
pd.testing.assert_frame_equal(sc.record_data_frame, expected_df, check_like=True)
# No zones for time 1.
z[:] = 0
sc.run_one_step(1)
se.run_one_step(1)
np.testing.assert_equal(len(sc.zones), 0)
np.testing.assert_equal(zone._conn_type, zn.Connection.ONE_TO_NONE)
expected_df = pd.DataFrame(
{
"time": [0, 1],
"zones": [1, 0],
"fragmentations": [np.nan, 0],
"captures": [np.nan, 0],
"area_captured_sum": [np.nan, 0],
"area_captured_max": [np.nan, 0],
}
)
pd.testing.assert_frame_equal(sc.record_data_frame, expected_df, check_like=True)
np.testing.assert_equal(se.record_data_frame.taxa.sum(), 1)
def test_neighborhood_structure(zone_example_grid):
mg, z = zone_example_grid
z[[10, 16]] = 1
sc = ZoneController(mg, zone_func)
np.testing.assert_equal(len(sc.zones), 1)
sc = ZoneController(mg, zone_func, neighborhood_structure="D4")
np.testing.assert_equal(len(sc.zones), 2)
np.testing.assert_raises(
ValueError, ZoneController, mg, zone_func, neighborhood_structure="D"
)
def test_many_to_one(zone_example_grid):
mg, z = zone_example_grid
# Create two zones for time 0.
z[[8, 9, 10, 12]] = 1
se = SpeciesEvolver(mg)
sc = ZoneController(mg, zone_func)
taxa = sc.populate_zones_uniformly(1)
se.track_taxa(taxa)
expected_df = pd.DataFrame(
{
"time": [0],
"zones": [2],
"fragmentations": [np.nan],
"captures": [np.nan],
"area_captured_sum": [np.nan],
"area_captured_max": [np.nan],
}
)
pd.testing.assert_frame_equal(sc.record_data_frame, expected_df, check_like=True)
np.testing.assert_equal(len(se.get_extant_taxon_objects(time=0)), 2)
# Modify elevation such that two zones each overlap the original two zones.
z[11] = 1
sc.run_one_step(1)
se.run_one_step(1)
np.testing.assert_equal(len(sc.zones), 1)
np.testing.assert_equal(sc.zones[0]._conn_type, zn.Connection.MANY_TO_ONE)
expected_df = pd.DataFrame(
{
"time": [0, 1],
"zones": [2, 1],
"fragmentations": [np.nan, 0],
"captures": [np.nan, 1],
"area_captured_sum": [np.nan, 12],
"area_captured_max": [np.nan, 12],
}
)
pd.testing.assert_frame_equal(sc.record_data_frame, expected_df, check_like=True)
np.testing.assert_equal(len(se.get_extant_taxon_objects(time=1)), 2)
def test_none_to_one(zone_example_grid):
mg, z = zone_example_grid
# No zones exist at time 0.
se = SpeciesEvolver(mg)
sc = ZoneController(mg, zone_func)
taxa = sc.populate_zones_uniformly(1)
se.track_taxa(taxa)
np.testing.assert_equal(len(sc.zones), 0)
expected_df = pd.DataFrame({
'time': [0],
'zones': [0],
'fragmentations': [np.nan],
'captures': [np.nan],
'area_captured_sum': [np.nan],
'area_captured_max': [np.nan]
})
pd.testing.assert_frame_equal(sc.record_data_frame,
expected_df,
check_like=True)
# Create a zone for time 1.
z[[9, 10, 11, 12]] = 1
sc.run_one_step(1)
se.run_one_step(1)
np.testing.assert_equal(len(sc.zones), 1)
np.testing.assert_equal(sc.zones[0]._conn_type, zn.Connection.NONE_TO_ONE)
expected_df = pd.DataFrame({
'time': [0, 1],
'zones': [0, 1],
'fragmentations': [np.nan, 0],
'captures': [np.nan, 0],
'area_captured_sum': [np.nan, 0],
'area_captured_max': [np.nan, 0]
})
pd.testing.assert_frame_equal(sc.record_data_frame,
expected_df,
check_like=True)
np.testing.assert_equal(se.record_data_frame.taxa.sum(), 0)
def test_min_area(zone_example_grid):
mg, z = zone_example_grid
# Create a zone for time 0.
z[[9, 11, 12]] = 1
sc = ZoneController(mg, zone_func, minimum_area=5)
np.testing.assert_equal(len(sc.zones), 1)
def test_pseudoextinction(zone_example_grid):
mg, z = zone_example_grid
z[[9, 10, 11, 12]] = 1
se = SpeciesEvolver(mg)
sc = ZoneController(mg, zone_func)
taxa = sc.populate_zones_uniformly(1, persists_post_speciation=False)
se.track_taxa(taxa)
z[11] = 0
sc.run_one_step(1)
se.run_one_step(1)
expected_df = pd.DataFrame(
{
"pid": [np.nan, 0, 0],
"type": 3 * [ZoneTaxon.__name__],
"t_first": [0, 1, 1],
"t_final": [1, np.nan, np.nan],
},
index=[0, 1, 2],
)
expected_df.index.name = "tid"
expected_df["pid"] = expected_df["pid"].astype("Int64")
expected_df["t_final"] = expected_df["t_final"].astype("Int64")
pd.testing.assert_frame_equal(se.taxa_data_frame, expected_df, check_like=True)
expected_df = pd.DataFrame(
{
"time": [0, 1],
"taxa": [1, 2],
"speciations": [np.nan, 2],
"extinctions": [np.nan, 0],
"pseudoextinctions": [np.nan, 1],
}
)
pd.testing.assert_frame_equal(se.record_data_frame, expected_df, check_like=True)
def test_one_to_many_to_one(zone_example_grid):
mg, z = zone_example_grid
z[[9, 10, 11, 12]] = 1
se = SpeciesEvolver(mg)
sc = ZoneController(mg, zone_func)
taxa = sc.populate_zones_uniformly(1, time_to_allopatric_speciation=1)
se.track_taxa(taxa)
z[11] = 0
sc.run_one_step(1)
se.run_one_step(1)
z[11] = 1
sc.run_one_step(1)
se.run_one_step(1)
np.testing.assert_equal(len(se.get_extant_taxon_objects()), 1)
