#creating regex for state names
for i in range(1, 3):
var_edit = "regex_state_edit_" + str(i)
var = "regex_state_" + str(i)
vars()[var] = ""
vars()[var_edit] = ""
with open('states_' + str(i) + '.txt', 'r') as states:
for state in states:
state = re.sub(r"\(.*\)", " ", state)
state_edit = ""
state = re.sub("\n", "", state)
vars()[var] += r"\b" + state + r"\b|"
if len(str(state)) <= 4:
continue
for state_edit in spellMistakes(state):
if state_edit == "Nagar" or len(state_edit) <= 4:
continue
vars()[var_edit] += r"\b" + state_edit + r"\b|"
vars()[var] = vars()[var][:len(vars()[var]) - 1]
vars()[var_edit] = vars()[var_edit][:len(vars()[var_edit]) - 1]
#location of test file
test_file_loc = "testAddress.json"
with open(test_file_loc, 'r') as file:
all_add = json.load(file)
#declaring ml variables
#maximum no of predictions made for a given address
no_of_predictions = 5
for i in range(1, 7):
var_edit = "regex_city_edit_" + str(i)
var = "regex_city_" + str(i)
vars()[var] = ""
vars()[var_edit] = ""
with open('tier_cities\\tier' + str(i) + 'cities.txt', 'r') as cities:
for city in cities:
city = re.sub(r"\(.*\)", " ", city)
city_edit = ""
city = re.sub("\n", "", city)
if city == "Nagar":
continue
vars()[var] += r"\b" + city + r"\b|"
if len(str(city)) <= 4:
continue
for city_edit in spellMistakes(city):
if city_edit == "Nagar" or len(city_edit) <= 4:
continue
vars()[var_edit] += r"\b" + city_edit + r"\b|"
vars()[var] = vars()[var][:len(vars()[var]) - 1]
vars()[var_edit] = vars()[var_edit][:len(vars()[var_edit]) - 1]
index = -1
#looping through all addresses and predicting
for dictionary in all_add:
actual_city = dictionary["city"]
add = dictionary["address"]
#removing utf-16 encoding from the address
add = re.sub(r"\\u....", " ", add)
#creating regex for state names
for i in range(1,3) :
var_edit = "regex_state_edit_"+str(i)
var = "regex_state_"+str(i)
vars()[var] = ""
vars()[var_edit] = ""
with open('states_'+str(i)+'.txt','r') as states :
for state in states :
state = re.sub(r"\(.*\)"," ",state)
state_edit = ""
state = re.sub("\n","",state)
vars()[var] += r"\b" + state + r"\b|"
if len(str(state)) <= 4 :
continue
for state_edit in spellMistakes(state) :
if state_edit == "Nagar" or len(state_edit) <= 4:
continue
vars()[var_edit] += r"\b" + state_edit + r"\b|"
vars()[var] = vars()[var][:len(vars()[var])-1]
vars()[var_edit] = vars()[var_edit][:len(vars()[var_edit])-1]
#location of test file
test_file_loc = "trainAddress.json"
with open(test_file_loc,'r') as file :
all_add = json.load(file)
#declaring ml variables
#maximum no of predictions made for a given address
for i in range(1,7) :
var_edit = "regex_city_edit_"+str(i)
var = "regex_city_"+str(i)
vars()[var] = ""
vars()[var_edit] = ""
with open('tier_cities\\tier' + str(i) + 'cities.txt','r') as cities :
for city in cities :
city = re.sub(r"\(.*\)"," ",city)
city_edit = ""
city = re.sub("\n","",city)
if city == "Nagar" :
continue
vars()[var] += r"\b" + city + r"\b|"
if len(str(city)) <= 4 :
continue
for city_edit in spellMistakes(city) :
if city_edit == "Nagar" or len(city_edit) <= 4:
continue
vars()[var_edit] += r"\b" + city_edit + r"\b|"
vars()[var] = vars()[var][:len(vars()[var])-1]
vars()[var_edit] = vars()[var_edit][:len(vars()[var_edit])-1]
index = -1
#looping through all addresses and predicting
for dictionary in all_add :
actual_city = dictionary["city"]
add = dictionary["address"]
#removing utf-16 encoding from the address
add = re.sub(r"\\u...."," ",add)
