Revision history [back]

Ironically, the LabelledBinaryTree() command only lets you label the root. But you can apply a canonical labeling and then map those labels to the desired ones.

import re
words = re.compile('\w+')

# labelled binary tree from string
def lbt_from_str(lbstr):
    labels = [Integer(w) for w in words.findall(lbstr)]
    blank = words.sub('', lbstr)
    work = LabelledBinaryTree(blank).canonical_labelling()
    wstr = '%s'%work
    canon = [Integer(w) for w in words.findall(wstr)]
    fixit = dict((canon[i] , labels[i]) for i in range(len(canon)))
    return work.map_labels(lambda z: fixit[z])
# end lbt_from_str

This doesn't check if the string has a valid structure, and it expects all nodes to be labeled (you should use 'None' for unlabeled ones). Please feel free to improve it.

Testing:

t = Permutation([1, 3, 2, 6, 5, 7, 4]).increasing_tree()
print t
show(t)
s = '%s'%t
test = lbt_from_str(s)
print test
show(test)

Ironically, the LabelledBinaryTree() command only lets you label the root. But you can apply a canonical labeling and then map those labels to the desired ones.

import re
words = re.compile('\w+')

# labelled binary tree from string
def lbt_from_str(lbstr):
    labels = [Integer(w) for w in words.findall(lbstr)]
    blank = words.sub('', lbstr)
    work = LabelledBinaryTree(blank).canonical_labelling()
    wstr = '%s'%work
    canon = [Integer(w) for w in words.findall(wstr)]
    fixit = dict((canon[i] , labels[i]) for i in range(len(canon)))
    return work.map_labels(lambda z: fixit[z])
# end lbt_from_str

This doesn't check if the string has a valid structure, and it expects all nodes to be labeled (you should use 'None' for unlabeled ones). with integers. *I will fix that shortly ... Please feel free to improve it.

Testing:

t = Permutation([1, 3, 2, 6, 5, 7, 4]).increasing_tree()
print t
show(t)
s = '%s'%t
test = lbt_from_str(s)
print test
show(test)

Ironically, the LabelledBinaryTree() command only lets you label the root. But you can apply a canonical labeling and then map those labels to the desired ones.

import re
words = re.compile('\w+')

# keep int labels as ints
def safeInt(word):
    try:
        return Integer(word)
    except TypeError:
        return word

# labelled binary tree from string
def lbt_from_str(lbstr):
    labels = [Integer(w) for w in words.findall(lbstr)]
    blank = words.sub('', lbstr)
    work = LabelledBinaryTree(blank).canonical_labelling()
    wstr = '%s'%work
    canon = [Integer(w) for w in words.findall(wstr)]
    labels = [safeInt(w) for w in words.findall(lbstr)]
    fixit = dict((canon[i] , dict((canon[i], labels[i]) for i in range(len(canon)))
    return work.map_labels(lambda z: fixit[z])
# end lbt_from_str

This doesn't check if the string has a valid structure, and it expects all nodes to be labeled have labels.

You could replace ' [' with integers. *I will fix that shortly ... Please feel free to improve it.'None['

Testing:

t = Permutation([1, 3, 2, 6, 5, 7, 4]).increasing_tree()
print t
show(t)
s = '%s'%t
test = lbt_from_str(s)
print test
show(test)