def test_bad_sampling(self):
"""
Test bad sampling simulation on an existing tree.
"""
tree = Tree(_TREES[-1])
ngesh.add_characters(
tree,
10,
k=4.0,
th=1.0,
z=1.045,
k_hgt=2.0,
th_hgt=1.1,
seed="myseed",
)
previous = tree.write()
ngesh.simulate_bad_sampling(tree, 0.5, seed="myseed")
digest = hashlib.sha256(
str(ngesh.tree2wordlist(tree)).encode("utf-8")).digest()
assert tree.write() != previous
assert (
digest ==
b'\x8e\xe3\x9fzN\xbe0\xaa\xe2a\xc5\x854\x87>\xe6"s1?\xc1\x08YqM\xc4\xdd9Zh\xb37'
)
def test_tree_output(self):
# Add characters to all test trees
trees = [
ngesh.add_characters(Tree(newick),
100,
k=4.0,
th=1.0,
e=1.05,
seed="myseed") for newick in _TREES
]
# Assert the first one
digest_nx = hashlib.sha256(
str(ngesh.tree2nexus(trees[0])).encode("utf-8")).digest()
digest_wl = hashlib.sha256(
str(ngesh.tree2wordlist(trees[0])).encode("utf-8")).digest()
assert (
digest_nx ==
b"E\xf8\x97\xb6*\x7f\xf4_j\x89\x02dn\x1d\xbe\xb0\xb6\xcd\xd9.\xca:\x9ft\xe2m\xc5y\xa5\xaa\x0fa"
)
assert (
digest_wl ==
b"\xb6\xc8i\xf4!\xf4l\x91\xb9\x8d\xb5Kae\x1aF\x9c\xfd\n \x06\xf2D\x1e<\xdd(U]6(\xbf"
)
# Test output for a tree without characters
tree_nochar = Tree(_TREES[0])
digest_nx_nochar = hashlib.sha256(
str(ngesh.tree2nexus(tree_nochar)).encode("utf-8")).digest()
assert (
digest_nx_nochar ==
b"A\x06N\x97n1iD\x01n\x07T\x99al\xa8)m\n\x93\x9ajKp\x07\x9bc\x8a\x03\x1a\x8d\xb7"
)
def tree_test():
CONCEPTS = 10
# generate a fixed tree for the tests
tree = ngesh.gen_tree(1.0, 0.5, max_time=1.0, labels="human", seed=13)
# TODO: need to add seed to `add_characters`
# TODO: need to sort the taxa in output in `tree2nexus`
tree = ngesh.add_characters(tree, CONCEPTS, 3.0, 0.5, seed=13)
# collect all taxa and their characters, provided the characters exist
# TODO: reuse code? already in tree2nexus
data = {leaf.name: leaf.chars for leaf in tree.get_leaves()}
# get the number of words to generate -- we generate all those that
# will be used at the beginning, following their path as if they
# always existed
# TODO: generate based on phonology of current available ones?
num_words = max([max(chars) for chars in data.values()]) + 1
words = abzu.kiss.random_words(num_words, param={})
print(tree)
print(ngesh.tree2nexus(tree))
print(data)
print(words, len(words))
for node in tree.iter_descendants():
print([node.name], node.dist)
print(dir(node))
def test_add_characters_with_hgt(self):
tree_hgt = Tree(_TREES[0])
ngesh.add_characters(
tree_hgt,
10,
k=4.0,
th=1.0,
z=1.045,
k_hgt=2.0,
th_hgt=1.1,
seed="myseed",
)
digest = hashlib.sha256(
str(ngesh.tree2wordlist(tree_hgt)).encode("utf-8")).digest()
assert (
digest ==
b"\xc0\x84\x8a\xa8\x0b\xd6\xf8\x1b\x9c\xc7\xe6\xaf]\xe7\xee\x105\x989\x9b\xcfke\xe5\xf3\x03\xdc\x17U\x14 \xac"
)
def new_tree(args):
"""
Generates and returns a new tree.
This function is a simple wrapper on the main `gen_tree()` and
`add_characters()` functions. It is intented for command-line usage,
but it can be used in code for quick prototyping.
Parameters
----------
args : Namespace
A namespace with the parameters for tree generation.
Returns
-------
tree: ete3 object
The randomly generated tree.
"""
# Generate the random tree
tree = ngesh.gen_tree(
args.birth,
args.death,
min_leaves=args.min_leaves,
max_time=args.max_time,
labels=args.labels,
lam=args.lam,
seed=args.seed,
)
# Add characters if requested
if args.num_chars:
tree = ngesh.add_characters(
tree,
args.num_chars,
args.k_mut,
args.th_mut,
k_hgt=args.k_hgt,
th_hgt=args.th_hgt,
e=args.e_mut,
seed=args.seed,
)
# Simulate bad sampling if requested
if args.sampling:
ngesh.simulate_bad_sampling(tree, args.sampling, seed=args.seed)
return tree
def test_add_characters(self):
# gamma parameters
NUM_CONCEPTS = 10
k = 4.0 # shape
th = 1.0 # scale
z = 1.045 # "zipf" correction
# Add characters to all trees, for coverage
trees = [
ngesh.add_characters(Tree(newick),
NUM_CONCEPTS,
k=k,
th=th,
z=z,
seed="myseed") for newick in _TREES
]
# Assert the first one
digest = hashlib.sha256(
str(ngesh.tree2wordlist(trees[0])).encode("utf-8")).digest()
assert (
digest ==
b"\xdc<\x1f\x10N\xbf\xcc\xc4|l26\x10\xfc\xbaN\xb7\\c\x8bB\xca\x95.\xcbH\x82T\xa3\xbd\xff\x15"
)