def newick_to_TB(s, name_type=None):
"""
Vertex names are used as vertices.
This is mostly for testing.
@param s: newick string
@return: undirected topology, branch lengths
"""
T, B, N = newick_to_TBN(s)
N = get_validated_name_map(N, name_type)
T = set(mkedge(N[a], N[b]) for a, b in T)
B = dict((mkedge(N[a], N[b]), x) for (a, b), x in B.items())
return T, B
def _BNv_to_newick(v_to_source, v_to_sinks, B, N, v):
"""
This is part of writing a newick string with branch lengths.
Note that this function does not add the semicolon termination.
@param v_to_source: a map from a vertex to its source
@param v_to_sinks: a map from a vertex to a set of sinks
@param B: branch lengths
@param N: map from vertices to names
@param v: a subtree root vertex
@return: a chunk of a newick string
"""
if v in v_to_source:
# a vertex that has a source should record its distance to its source
blen = B[mkedge(v, v_to_source[v])]
suffix = ':' + str(blen)
else:
suffix = ''
if v not in v_to_sinks:
if v not in N:
raise ValueError('expected leaf vertices to be named')
return str(N[v]) + suffix
sinks = sorted(v_to_sinks[v])
arr = [_BNv_to_newick(v_to_source, v_to_sinks, B, N, x) for x in sinks]
internal_vertex_name = str(N.get(v, ''))
return '(' + ', '.join(arr) + ')' + internal_vertex_name + suffix
def set_branch_length(self, v, blen):
"""
Note that a branch length can be set to a root.
This happens during the construction of the tree
when the subtree has not yet been connected to the rest of the tree.
"""
if v in self.v_to_source:
edge = mkedge(v, self.v_to_source[v])
self.B[edge] = blen
else:
self.v_to_hanging_length[v] = blen
def newick_to_T(s, name_type=None):
"""
Everything to do with branch lengths is ignored.
Vertex names are used as vertices.
This is mostly for testing.
@param s: newick string
@return: undirected topology
"""
T, N = newick_to_TN(s)
N = get_validated_name_map(N, name_type)
T = set(mkedge(N[a], N[b]) for a, b in T)
return T
def _Bv_to_newick(v_to_source, v_to_sinks, B, v):
"""
This is part of writing a newick string with branch lengths.
Note that this function does not add the semicolon termination.
@param v_to_source: a map from a vertex to its source
@param v_to_sinks: a map from a vertex to a set of sinks
@param B: branch lengths
@param v: a subtree root vertex
@return: a chunk of a newick string
"""
if v in v_to_source:
# a vertex that has a source should record its distance to its source
blen = B[mkedge(v, v_to_source[v])]
suffix = ':' + str(blen)
else:
suffix = ''
if v not in v_to_sinks:
return str(v) + suffix
sinks = sorted(v_to_sinks[v])
arr = [_Bv_to_newick(v_to_source, v_to_sinks, B, x) for x in sinks]
return '(' + ', '.join(arr) + ')' + str(v) + suffix
if name_type:
N = dict((v, name_type(n)) for v, n in N.items())
nvertices = len(N)
names = N.values()
if any(n is None for n in names):
raise ValueError(
'expected a name for each vertex, including internal vertices')
if len(set(names)) < nvertices:
raise ValueError('expected unique vertex names')
return N
# Testing
g_example_T = set([
mkedge(2,1),
mkedge(2,3),
mkedge(2,4),
mkedge(3,5),
mkedge(3,6),
mkedge(6,7)])
g_example_B = {
mkedge(2,1) : 1,
mkedge(2,3) : 2,
mkedge(2,4) : 2,
mkedge(3,5) : 3,
mkedge(3,6) : 3,
mkedge(6,7) : 3}
class TestFtreeIO(unittest.TestCase):