def __call__(self, T, leafRanks=None):
self.__init__()
# to indicate whether we are in the 2nd pass
self.setLeafRanks = leafRanks
T.dfs_traverse(self)
lca = LCA.LCA(self.pa)
return (lca, self.nodes, self.getLeafRanks)
def test_LCA():
assert LCA.LCA(LCA.testTree, 8, 9).value == 2
assert LCA.LCA(LCA.testTree, 10, 3).value == 1
assert LCA.LCA(LCA.testTree, 5, 6).value == 5
assert LCA.LCA(LCA.testTree, 1, 6).value == 1
assert LCA.LCA(LCA.testTree, 9, 6).value == 1
assert LCA.LCA(LCA.testTree, 4, 6).value == 3
def test_LCA_sameValue(self):
self.assertEqual("A", LCA.LCA(LCA.graph, "A", "A"))
self.assertEqual("D", LCA.LCA(LCA.graph, "D", "D"))
def test_LCA_invalidValue(self):
self.assertEqual("", LCA.LCA(LCA.graph, "Z", "X"))
def test_LCA_normal(self):
self.assertEqual("A", LCA.LCA(LCA.graph, "A", "B"))
self.assertEqual("B", LCA.LCA(LCA.graph, "C", "D"))
def test_none_node_LCA():
test = LCA.Node(None)
assert LCA.LCA(test, None, None).value == None
assert LCA.LCA(test, None, 1) == -1
assert LCA.LCA(test, 1, None) == -1
assert LCA.LCA(test, 1, 1) == -1
def test_none_LCA():
assert LCA.LCA(LCA.testTree, None, 1) == -1
assert LCA.LCA(LCA.testTree, 1, None) == -1
assert LCA.LCA(LCA.testTree, None, None) == -1
def test_LCA():
assert LCA.LCA(LCA.testTree, 10, 16) == -1
assert LCA.LCA(LCA.testTree, 22, 0) == -1
assert LCA.LCA(LCA.testTree, 22, 2) == -1
def build_LCA(G, S, verbose=False):
"""
build the LCA mapping from given G to given S:
1. preprocess S
2. construct leaf-mapping
3. DFS through G, map G to S
"""
class DFS_species_preLCA(newick.tree.TreeVisitor):
"""
preprocess S for LCA:
1. build the child-parent dictionary for LCA
2. build the leaf-node dictionary
"""
def __init__(self):
self.Pa = {} # child-parent relation, Pa()
self.Le = {} # 'id'-leaf relation, Le('id')
self.nodes = []
def pre_visit_edge(self, src, bootstrap, length, dst):
self.Pa.setdefault(dst, src)
def post_visit_tree(self, t):
self.nodes.append(t)
def visit_leaf(self, s):
id = s.leaves_ids[0]
if self.Le.has_key(id):
raise "ERROR: Species tree has duplicated taxon: %s" % id
self.Le.setdefault(s.leaves_ids[0], s)
self.nodes.append(s)
class DFS_gene_LCA(newick.tree.TreeVisitor):
"""
recursively build up the LCA mapping
"""
def __init__(self, Ls, LCAs):
self.Ls = Ls # leaves of S, Ls['id'] = leave
self.LCAs = LCAs # LCA of S
self.nodes = []
def pre_visit_tree(self, t):
t.M = None
def post_visit_tree(self, t):
self.nodes.append(t)
def post_visit_edge(self, src, bo, le, trg):
if not trg.M:
raise "ERROR: Gene tree failed LCA mapping at: %s" % trg.leaves
src.M = self.LCAs(src.M, trg.M)
def visit_leaf(self, g):
if hasattr(g, "interleaf"):
g.M = self.LCAs(*g.interleaf)
else:
id = g.leaves_ids[0]
if not self.Ls.has_key(id):
raise "ERROR: Gene tree has orphan taxon: %s" % id
g.M = self.Ls[id]
self.nodes.append(g)
# step 1: preprocess S
dfsS = DFS_species_preLCA()
S.dfs_traverse(dfsS)
lcaS = LCA.LCA(dfsS.Pa)
# test:
#print dfsS.Pa
#print dfsS.Le
#print dfsS.Le['a'], dfsS.Le['b']
#print lcaS(dfsS.Le['a'], dfsS.Le['b'])
# step 2: build the leaf-mapping
# step 3: build the LCA-mapping
dfsG = DFS_gene_LCA(dfsS.Le, lcaS)
G.dfs_traverse(dfsG)
# test:
# a little short-cut for testing purpose
if verbose:
S.nodes = dfsS.nodes
G.nodes = dfsG.nodes