def generate_single_labeled_history(tree,
pop_size=None,
rng=None):
# get collection of nodes organized by the order
# of coalescence as given by the level
node_event_order = {}
event_nodes = {}
num_leaves = 0
for nd in tree.postorder_node_iter():
if nd.is_leaf():
node_event_order[nd] = 0
num_leaves += 1
else:
eo = max([node_event_order[ch] for ch in nd.child_nodes()]) + 1
node_event_order[nd] = eo
try:
event_nodes[node_event_order[nd]].append(nd)
except KeyError:
event_nodes[node_event_order[nd]] = [nd]
# set the node ages
node_ages = {}
event_idx_list = sorted(event_nodes.keys())
for nd in event_nodes[0]:
node_ages[nd] = 0.0
current_tree_max_age = 0.0
num_nodes_remaining = num_leaves
for event_idx in event_idx_list[1:]:
for nd in event_nodes[event_idx]:
child_nodes = nd.child_nodes()
max_ch_age = max([node_ages[ch] for ch in child_nodes])
min_age = max(current_tree_max_age, max_ch_age)
if rng is None:
# rng is not given: use expected waiting time
wt = coalescent.expected_tmrca(
n_genes=num_nodes_remaining,
pop_size=pop_size,
n_to_coalesce=len(child_nodes))
else:
# rng is given: draw random waiting time
wt = coalescent.time_to_coalescence(
n_genes=num_nodes_remaining,
pop_size=pop_size,
n_to_coalesce=len(child_nodes))
node_ages[nd] = min_age + wt
current_tree_max_age += wt
num_nodes_remaining = num_nodes_remaining - len(child_nodes) + 1
# set the edge lengths according to the node ages
for nd in tree.preorder_node_iter():
if nd.parent_node is None:
nd.edge.length = 0.0
else:
nd.edge.length = node_ages[nd.parent_node] - node_ages[nd]
return tree
def generate_single_labeled_history(tree, pop_size=None, rng=None):
# get collection of nodes organized by the order
# of coalescence as given by the level
node_event_order = {}
event_nodes = {}
num_leaves = 0
for nd in tree.postorder_node_iter():
if nd.is_leaf():
node_event_order[nd] = 0
num_leaves += 1
else:
eo = max([node_event_order[ch] for ch in nd.child_nodes()]) + 1
node_event_order[nd] = eo
try:
event_nodes[node_event_order[nd]].append(nd)
except KeyError:
event_nodes[node_event_order[nd]] = [nd]
# set the node ages
node_ages = {}
event_idx_list = sorted(event_nodes.keys())
for nd in event_nodes[0]:
node_ages[nd] = 0.0
current_tree_max_age = 0.0
num_nodes_remaining = num_leaves
for event_idx in event_idx_list[1:]:
for nd in event_nodes[event_idx]:
child_nodes = nd.child_nodes()
max_ch_age = max([node_ages[ch] for ch in child_nodes])
min_age = max(current_tree_max_age, max_ch_age)
if rng is None:
# rng is not given: use expected waiting time
wt = coalescent.expected_tmrca(n_genes=num_nodes_remaining,
pop_size=pop_size,
n_to_coalesce=len(child_nodes))
else:
# rng is given: draw random waiting time
wt = coalescent.time_to_coalescence(
n_genes=num_nodes_remaining,
pop_size=pop_size,
n_to_coalesce=len(child_nodes))
node_ages[nd] = min_age + wt
current_tree_max_age += wt
num_nodes_remaining = num_nodes_remaining - len(child_nodes) + 1
# set the edge lengths according to the node ages
for nd in tree.preorder_node_iter():
if nd.parent_node is None:
nd.edge.length = 0.0
else:
nd.edge.length = node_ages[nd.parent_node] - node_ages[nd]
return tree
def _generate_labeled_history(tree,
coal_event_list,
num_leaves=None,
pop_size=None,
rng=None):
# set the node ages
node_ages = {}
for nd in coal_event_list[0]:
node_ages[nd] = 0.0
current_tree_max_age = 0.0
if num_leaves is None:
num_nodes_remaining = num_leaves
else:
num_nodes_remaining = len([lf for lf in tree.leaf_iter()])
# process events in order of occurrence
for event_nodes in coal_event_list[1:]:
for nd in event_nodes:
child_nodes = nd.child_nodes()
max_ch_age = max([node_ages[ch] for ch in child_nodes])
min_age = max(current_tree_max_age, max_ch_age)
if rng is None:
# rng is not given: use expected waiting time
wt = coalescent.expected_tmrca(
n_genes=num_nodes_remaining,
pop_size=pop_size,
n_to_coalesce=len(child_nodes))
else:
# rng is given: draw random waiting time
wt = coalescent.time_to_coalescence(
n_genes=num_nodes_remaining,
pop_size=pop_size,
n_to_coalesce=len(child_nodes))
node_ages[nd] = min_age + wt
current_tree_max_age += wt
num_nodes_remaining = num_nodes_remaining - len(child_nodes) + 1
# set the edge lengths according to the node ages
for nd in tree.preorder_node_iter():
if nd.parent_node is None:
nd.edge.length = 0.0
else:
nd.edge.length = node_ages[nd.parent_node] - node_ages[nd]
return tree
def _generate_labeled_history(tree,
coal_event_list,
num_leaves=None,
pop_size=None,
rng=None):
# set the node ages
node_ages = {}
for nd in coal_event_list[0]:
node_ages[nd] = 0.0
current_tree_max_age = 0.0
if num_leaves is None:
num_nodes_remaining = num_leaves
else:
num_nodes_remaining = len([lf for lf in tree.leaf_iter()])
# process events in order of occurrence
for event_nodes in coal_event_list[1:]:
for nd in event_nodes:
child_nodes = nd.child_nodes()
max_ch_age = max([node_ages[ch] for ch in child_nodes])
min_age = max(current_tree_max_age, max_ch_age)
if rng is None:
# rng is not given: use expected waiting time
wt = coalescent.expected_tmrca(n_genes=num_nodes_remaining,
pop_size=pop_size,
n_to_coalesce=len(child_nodes))
else:
# rng is given: draw random waiting time
wt = coalescent.time_to_coalescence(
n_genes=num_nodes_remaining,
pop_size=pop_size,
n_to_coalesce=len(child_nodes))
node_ages[nd] = min_age + wt
current_tree_max_age += wt
num_nodes_remaining = num_nodes_remaining - len(child_nodes) + 1
# set the edge lengths according to the node ages
for nd in tree.preorder_node_iter():
if nd.parent_node is None:
nd.edge.length = 0.0
else:
nd.edge.length = node_ages[nd.parent_node] - node_ages[nd]
return tree
def main():
"""
Main CLI handler.
"""
parser = argparse.ArgumentParser()
parser.add_argument("-f",
"--format",
dest="schema",
type=str,
default="nexus",
choices=["nexus", "newick"],
help="input data format (default='%(default)s')")
parser.add_argument("-l",
"--label",
type=str,
default="[TESTING]",
help="label for test")
parser.add_argument(
"-r",
"--regime",
dest="regime",
type=str,
default="mean",
choices=[
"mean-coalescent", "converse-coalescent", "uniform-coalescent"
],
help=
"expected regime under which the coalescent intervals were generated (default='%(default)s')"
)
parser.add_argument("-p",
"--precision",
type=float,
default=0.01,
help="numerical precision (default=%(default)s)")
args = parser.parse_args()
fails = 0
trees = dendropy.TreeList.get_from_stream(sys.stdin, schema=args.schema)
for tidx, tree in enumerate(trees):
tree_fails = 0
# check structure
if not tree._debug_tree_is_valid():
sys.stderr.write("{} Tree {}: structure is invalid".format(
args.label, tidx + 1))
tree_fails += 1
# ensure parent ages > child ages
tree.calc_node_ages(check_prec=args.precision)
for nd in tree.postorder_node_iter():
if nd.parent_node is not None:
if nd.age > nd.parent_node.age:
sys.stderr.write(
"{} Tree {}: Node '{}': age ({}) is greater than parent age ({})"
.format(args.label, tidx + 1, nd.label, nd.age,
nd.parent_node.age))
tree_fails += 1
# check waiting times
num_tips = len(tree.leaf_nodes())
wf = coalescent.extract_coalescent_frames(
tree, check_ultrametricity_prec=args.precision)
num_lineages = sorted(wf.keys())
for n in num_lineages:
wt = wf[n]
if args.regime == "mean-coalescent":
exp_wt = coalescent.expected_tmrca(n)
elif args.regime == "converse-coalescent":
exp_wt = coalescent.expected_tmrca(num_tips - n + 2)
elif args.regime == "uniform-coalescent":
exp_wt = 1.0
else:
raise RuntimeError("Unsupported regime: '{}'".format(
args.regime))
if abs(exp_wt - wt) > args.precision:
sys.stderr.write(
"{} Tree {}: Waiting time for coalescence event with {} lineages: expecting {} but found {}\n"
.format(args.label, tidx + 1, n, exp_wt, wt))
tree_fails += 1
if tree_fails > 0:
sys.stderr.write("{} Tree {} failed: {}\n".format(
args.label, tidx + 1, tree.compose_newick()))
fails += tree_fails
if fails > 0:
sys.exit(1)
else:
sys.exit(0)
def main():
"""
Main CLI handler.
"""
parser = argparse.ArgumentParser()
parser.add_argument("-f", "--format",
dest="schema",
type=str,
default="nexus",
choices=["nexus", "newick"],
help="input data format (default='%(default)s')")
parser.add_argument("-l", "--label",
type=str,
default="[TESTING]",
help="label for test")
parser.add_argument("-r", "--regime",
dest="regime",
type=str,
default="mean",
choices=["mean-coalescent", "converse-coalescent", "uniform-coalescent"],
help="expected regime under which the coalescent intervals were generated (default='%(default)s')")
parser.add_argument("-p", "--precision",
type=float,
default=0.01,
help="numerical precision (default=%(default)s)")
args = parser.parse_args()
fails = 0
trees = dendropy.TreeList.get_from_stream(sys.stdin, schema=args.schema)
for tidx, tree in enumerate(trees):
tree_fails = 0
# check structure
if not tree._debug_tree_is_valid():
sys.stderr.write("{} Tree {}: structure is invalid".format(args.label, tidx+1))
tree_fails += 1
# ensure parent ages > child ages
tree.calc_node_ages(check_prec=args.precision)
for nd in tree.postorder_node_iter():
if nd.parent_node is not None:
if nd.age > nd.parent_node.age:
sys.stderr.write("{} Tree {}: Node '{}': age ({}) is greater than parent age ({})".format(
args.label, tidx+1, nd.label, nd.age, nd.parent_node.age))
tree_fails += 1
# check waiting times
num_tips = len(tree.leaf_nodes())
wf = coalescent.extract_coalescent_frames(tree, check_ultrametricity_prec=args.precision)
num_lineages = sorted(wf.keys())
for n in num_lineages:
wt = wf[n]
if args.regime == "mean-coalescent":
exp_wt = coalescent.expected_tmrca(n)
elif args.regime == "converse-coalescent":
exp_wt = coalescent.expected_tmrca(num_tips - n + 2)
elif args.regime == "uniform-coalescent":
exp_wt = 1.0
else:
raise RuntimeError("Unsupported regime: '{}'".format(args.regime))
if abs(exp_wt - wt) > args.precision:
sys.stderr.write("{} Tree {}: Waiting time for coalescence event with {} lineages: expecting {} but found {}\n".format(
args.label, tidx+1, n, exp_wt, wt))
tree_fails += 1
if tree_fails > 0:
sys.stderr.write("{} Tree {} failed: {}\n".format(args.label, tidx+1, tree.compose_newick()))
fails += tree_fails
if fails > 0:
sys.exit(1)
else:
sys.exit(0)
