def test_list_built(self):
bwt = BWT([1, 2])
self.assertEqual(bwt.table[0], [BWT.END, 2, 1])
self.assertEqual(bwt.transform(), [1, 2, BWT.END])
self.assertEqual(bwt.C(1), 1) # There is one 1 in the string
self.assertEqual(bwt.F(1, 1), 1) # There is one 1 in the last one characters
def test_alignment(self):
bwt = BWT(list("banana"))
# bwt.print_table()
string = "ana" # This can be a string, because it supports random access and slicing.
self.assertEqual(bwt.L(string), 3)
self.assertEqual(bwt.U(string), 4)
self.assertEqual(sorted(bwt.get_start_indices(string)), [3,5]) # return is index of last token in match
class MarkovModel:
def __init__(self, string):
self.string = string
self.bwt = BWT(self.string)
def get_n_tokens(self, context, n):
'''
Returns up to n tokens that follow the list of tokens given in context in
the source string, or None if no such tokens exist.
'''
indices = self.bwt.get_start_indices(context)
if indices == []:
return None
index = choice(indices)
return self.get_n_gram_at_index(n, index)
def get_n_gram_at_index(self, n, index):
return self.string[index+1:min(index+1+n ,len(self.string)-1)]
def get_all_possible_n_grams(self, context, n):
indices = self.bwt.get_start_indices(context)
return (self.get_n_gram_at_index(n, index) for index in indices)
class ApproximateMatcher:
def __init__(self, target):
self._text = target + '$'
self._bwt = BWT(self._text)
# return indices in target that contain
# matches of string pattern with up to d
# mismatches
def get_matches(self, pattern, d):
# initialze seed and check object
seed_checker = SeedChecker(pattern, d)
# for each seed k-mer in pattern
for seed, seed_index in seed_checker.enumerate():
# find exact matches of seed using BWT
indices = self._bwt.get_matches(seed)
# add candidate approximate matches based on
# seed exact matches
seed_checker.add_candidates(indices, seed_index)
# verify that candidate approximate matches are within
# minimum edit distance, and return final matches
matches = seed_checker.filter_candidates(self._text)
return matches
class ApproximateMatcher:
def __init__(self, target):
self._text = target + "$"
self._bwt = BWT(self._text)
# return indices in target that contain
# matches of string pattern with up to d
# mismatches
def get_matches(self, pattern, d):
# initialze seed and check object
seed_checker = SeedChecker(pattern, d)
# for each seed k-mer in pattern
for seed, seed_index in seed_checker.enumerate():
# find exact matches of seed using BWT
indices = self._bwt.get_matches(seed)
# add candidate approximate matches based on
# seed exact matches
seed_checker.add_candidates(indices, seed_index)
# verify that candidate approximate matches are within
# minimum edit distance, and return final matches
matches = seed_checker.filter_candidates(self._text)
return matches
def __init__(self, string): self.string = string self.bwt = BWT(self.string)
def __init__(self, target):
self._text = target + '$'
self._bwt = BWT(self._text)
def __init__(self, target):
self._text = target + "$"
self._bwt = BWT(self._text)