def memoized_leven_dist(s1, s2):
memory = lev_dist_cache[s1][s2]
if not (memory is None):
return memory
else:
output = leven_dist(s1, s2)
lev_dist_cache[s1][s2] = output
return output
def find_station(stat_n): # finds the bike station using the Levenshtein distance if necessary if stat_n in stations_hs: return stations_hs[stat_n] elif stat_n.strip() in stations_hs: return stations_hs[stat_n.strip()] elif stat_n in aliases: return stations_hs[aliases[stat_n]] else: print 'couldn\'t find "%s"' % stat_n return min(stations_hs.values(), key=lambda x: leven_dist(x, stat_n))
def find_station(stat_n):
# finds the bike station using the Levenshtein distance if necessary
if stat_n in stations_hs:
return stations_hs[stat_n]
elif stat_n.strip() in stations_hs:
return stations_hs[stat_n.strip()]
elif stat_n in aliases:
return stations_hs[aliases[stat_n]]
else:
print 'couldn\'t find "%s"' % stat_n
return min(stations_hs.values(), key=lambda x: leven_dist(x, stat_n))
def levenshtein_multi_char_inserts(s_target, s_list):
# Stores the action that resulted in the best score at a location in the string
dp_memo = [{'cost': 0, 'from': None, 'via': None} for _ in s_target]
for s_index, entry in enumerate(dp_memo):
if s_index == 0:
current_cost = 0
else:
current_cost = dp_memo[s_index - 1]['cost']
for s in s_list:
upperbound = int(min(len(s) * 1.5, len(s_target) - s_index))
for i in xrange(1, upperbound + 1):
substring_target = s_target[s_index:s_index + i]
s_cost = current_cost + 1 + leven_dist(s, substring_target)
if dp_memo[s_index + i -
1]['from'] is None or dp_memo[s_index + i -
1]['cost'] > s_cost:
dp_memo[s_index + i - 1]['cost'] = s_cost
dp_memo[s_index + i - 1]['from'] = s_index - 1
dp_memo[s_index + i - 1]['via'] = s
return dp_memo
def find_station_sub(stat_n):
# finds the bike station using the Levenshtein distance if necessary
if stat_n in subway_stats_name:
return subway_stats_name[stat_n]
else:
return min(subway_stats.values(), key=lambda x: leven_dist(x, stat_n))
def get_second_structural_distance(p1, p2):
w1 = p1.get_str_representation()
w2 = p2.get_str_representation()
return leven_dist(w1, w2)
def find_station_sub(stat_n): # finds the bike station using the Levenshtein distance if necessary if stat_n in subway_stats_name: return subway_stats_name[stat_n] else: return min(subway_stats.values(), key=lambda x: leven_dist(x, stat_n))
