def get_genderize(names, outfile):
genderize = Genderize(user_agent="Kevin_Bonham", api_key=genderize_key)
for i in range(0, len(fixed_names), 10):
query = fixed_names[i : i + 10]
genders = genderize.get(query)
names_dict = {}
for gender in genders:
n = gender["name"]
g = gender["gender"]
if g != None:
p = gender["probability"]
c = gender["count"]
else:
p = None
c = None
n = gender["name"]
g = gender["gender"]
names_dict[n] = {"gender": g, "probability": p, "count": c}
with open(outfile, "w+") as f:
f.write(json.dumps(names_dict, indent=4))
def prob(self, name, binary=False):
# guess method to check names dictionary
if (self.config['DEFAULT']['genderize'] == 'no'):
v = Genderize().get([name])
elif (self.config['DEFAULT']['genderize'] == 'yes'):
fichero = open(self.config['DEFAULT']['genderizefile'], "r+")
apikey = fichero.readline().rstrip()
v = Genderize(
user_agent='GenderizeDocs/0.0',
api_key=apikey).get([name])
prob = v[0]['probability']
return prob
def gender(json_dict, first_name):
"""
Get the gender of the patient
:param dict json_dict: full response of the vision api in dict
:param str first_name: first_name of the patients
:return gender in case of printed gender
"""
male_guess_list = ['M', 'Male', 'm']
female_guess_list = ['F', 'Female', 'f']
gender_printed = get_gender_if_printed(json_dict)
if gender_printed:
if gender_printed == 'm' or gender_printed == 'male':
return 'M'
elif gender_printed == 'f' or gender_printed == 'female':
return 'F'
title_gender = get_gender_using_title(json_dict, first_name)
if title_gender:
return title_gender
try:
li = Genderize().get([first_name])[0]
if li['gender'] == 'male':
if check_for_gender(json_dict, male_guess_list):
return "M"
elif li['gender'] == 'female':
if check_for_gender(json_dict, female_guess_list):
return "F"
return None
except Exception as e:
print("Genderize not responding {}".format(e))
return None
def test_more_than_10_names():
"""
Retrieve 20 names, which requires multiple HTTP requests
now that the API has a 10-name limit.
"""
names = [
"Emma",
"Olivia",
"Ava",
"Isabella",
"Sophia",
"Mia",
"Charlotte",
"Amelia",
"Evelyn",
"Abigail",
"Liam",
"Noah",
"William",
"James",
"Logan",
"Benjamin",
"Mason",
"Elijah",
"Oliver",
"Jacob",
]
response = Genderize().get(names)
assert len(names) == len(response)
for name, namedata in zip(names, response):
assert name == namedata['name'],\
'Expected names to be returned in same order'
def genderize(genderize_api_key):
""" Function to set up Genderize
"""
genderize_obj = Genderize(user_agent='GenderizeCommunities/0.0', \
api_key=genderize_api_key)
return genderize_obj
def __get_gender(names):
"""
TODO: documentation
"""
try:
genders = Genderize().get(names, country_id="IT")
male = 0
male_count = 0
female = 0
female_count = 0
for gender in genders:
if gender['gender'] != None:
if gender['gender'] == "female":
female_count += 1
female += gender['probability']
else:
male_count += 1
male += gender['probability']
if male > 0:
male = male / male_count
if female > 0:
female = female / female_count
if male > 0 or female > 0:
if female > male:
return " Donna "
else:
return " Uomo "
else:
return " PERSONA "
except:
print(names)
return " PERSONARICHIESTAFALLITA "
def get_author_genders():
app.logger.info(request.json)
authors = tuple(request.json)
# Fetch authors from DB
query = "select authors.name, authors.gender, authors.gender_source from authors where authors.name IN %s"
conn = getConnection()
cur = conn.cursor()
cur.execute(query, (authors, ))
rows = cur.fetchall()
genders = {row[0]: (row[1], row[2]) for row in rows}
# Find missing authors
toFetch = list(set([x for x in authors if x not in genders]))
fetched = []
repsonse = Genderize(api_key="a619730661a7ce6b4f8e8e6b047046a2").get(
[x.split(" ")[0] for x in toFetch])
for author, resp in zip(toFetch, repsonse):
fetched.append((author, resp["gender"], "genderize"))
# Add fetched to response
for author, gender, source in fetched:
genders[author] = (gender, source)
# Add fetched to DB
from psycopg2.extras import execute_values
execute_values(
cur, "INSERT INTO authors (name, gender, gender_source) VALUES %s",
fetched)
conn.close()
return genders
def test_integration_single():
"""
Retrieve a single name.
"""
expected = 'male'
actual = Genderize().get1('Peter')['gender']
assert expected == actual,\
"Expected {0}, got {1}".format(expected, actual)
def test_with_headers():
"""
Calls the API server with an invalid API key.
Should result in an exception.
"""
headers = None
result = Genderize().get('Peter', retheader=True)
headers = result['headers']
assert headers, "Expected response headers to be returned"
def main():
inLines = open(INFILE, mode='r', encoding='ISO-8859-1')
# create array of unknown gender characters
unknown = []
for line in inLines:
fields = line.rsplit('\t')
if fields[4] == '?':
name = fields[4].split(' ')
unknown.append(fields[1])
# creates Genderize and gets genders
genderize = Genderize(
user_agent='GenderizeDocs/0.0',
api_key=config.api_key,
timeout=60)
results = []
for name in unknown:
results.append(gender_genderizer(genderize, name))
# sets pointer to beginning of file
inLines.seek(0)
outLines = open(OUTFILE, mode='w', encoding='ISO-8859-1')
# counts
changed = 0
unchanged = 0
i = 0
for line in inLines:
fields = line.rsplit('\t')
name = fields[1]
if fields[4] == '?':
prob = results[i]['probability']
if prob > 0.6:
gender = results[i]['gender'][0:1]
print('CHANGED:\t%s -> %s, %.2f' % (name, gender, prob))
fields[4] = gender
changed += 1
else:
print('UNCHANGED:\t%s' % (name))
unchanged += 1
i += 1
# print updated character metadata
newline = '\t'.join(fields)
outLines.write(newline)
print('----------------------------------------')
print('TOTAL UNKNOWN:\t\t%d' % (len(unknown)))
print('TOTAL CHANGED:\t\t%d' % (changed))
print('TOTAL UNCHANGED:\t%d' % (unchanged))
outLines.close()
inLines.close()
def guess(self, name, binary=False):
# guess method to check names dictionary
if (self.config['DEFAULT']['genderize'] == 'no'):
v = Genderize().get([name])
elif (self.config['DEFAULT']['genderize'] == 'yes'):
fichero = open(self.config['DEFAULT']['genderizefile'], "r+")
apikey = fichero.readline().rstrip()
v = Genderize(
user_agent='GenderizeDocs/0.0',
api_key=apikey).get([name])
g = v[0]['gender']
if ((g == 'female') and binary):
guess = 0
elif ((g == 'male') and binary):
guess = 1
elif (not(binary)):
guess = g
return guess
def main():
# Creates year folders if they don't exist; clears them if they do
for year in range(1975, 2016):
clear_dir('%s/%d' % (DESTFOL, year))
genderize = Genderize(user_agent='GenderizeDocs/0.0',
api_key=config.api_key,
timeout=60)
femCount = 0
maleCount = 0
unkCount = 0
for year in range(1975, 2016):
print('Gendering movies in %d...' % (year))
movies = read_folder_dict('%s/%d' % (SRCFOL, year), year)
for movie in movies:
characters = {}
title = movie['title']
lines = movie['text'].split('\n')
outFile = open('%s/%d/%s' % (DESTFOL, year, title),
mode='w',
encoding='ISO-8859-1')
for i in range(2, len(lines)):
fields = lines[i].split('\t')
if len(fields) < 2:
continue
name = fields.pop(0)
gender = getGender(name, characters, genderize)
outFile.write('%s\t%s\t%s\n' %
(name, gender, ' '.join(fields)))
for gender in characters.values():
if gender == '?':
unkCount += 1
elif gender == 'f':
femCount += 1
elif gender == 'm':
maleCount += 1
outFile.close()
print('Finished %s...' % (title))
print('----------------------------------------')
print('NUMBER OF CHARACTERS')
print('\tMale:\t\t%d' % (maleCount))
print('\tFemale:\t\t%d' % (femCount))
print('\tUnknown:\t%d' % (unkCount))
def get_gender_using_genderize_api(name):
"""
:param a string, ideally the first name
:returns a dictionary of name, gender, % accuracy
Description:Given a name as a string, this function returns a dic with the
possible gender
NOTE: This uses the genderize api
"""
if name:
return Genderize().get(name)
def test_invalid_api_key():
"""
Calls the API server with an invalid API key.
Should result in an exception.
"""
caught = False
try:
Genderize(api_key='invalid_api_key').get1('Peter')
except GenderizeException:
caught = True
assert caught, "Expected a GenderizeException to be thrown"
async def gender(self, ctx, *, name: str):
list = [f"{name}"]
em = discord.Embed(title="Gender")
gender = Genderize().get(list)
result = gender[0]
em.add_field(name="Name", value=f'{result["name"]}', inline=True)
em.add_field(name='Gender', value=f'{result["gender"]}', inline=False)
em.add_field(name='Probability',
value=f'{result["probability"]}',
inline=True)
await ctx.send(embed=em)
def test_integration():
"""
Integration test from the readme. Calls the real Genderize.io API server.
"""
expected_genders = {
'James': 'male',
'Eva': 'female',
'Thunderhorse': None,
}
actual_genders = dict((elem['name'], elem['gender'])
for elem in Genderize().get(expected_genders.keys()))
assert expected_genders == actual_genders,\
"Expected {0}, got {1}".format(expected_genders, actual_genders)
def gender():
gende = []
prenom = ""
#On defini le prenom via une RegEx qui prend le premier mot du CV
defPrenom = re.findall('\A[a-zA-Z{Ë, Ï, Ö, Œ, ï, ö, é,œ,â, ë, ç, ô, -}]+ ',
txt)
#On supprime l'espace
for suppEsp in defPrenom:
prenom = suppEsp.strip()
#on defini le sexe a partir du prenom
sexe = Genderize().get1(prenom)
gende.append(sexe['gender'])
return gende
def test_with_headers():
"""
Retrieve a single name with response headers.
"""
response = Genderize().get(['Peter'], retheader=True)
assert response['data'][0][
'name'] == 'Peter', "Expected name data to be returned"
assert response['headers'], "Expected response headers to be returned"
for header in [
'X-Rate-Limit-Limit', 'X-Rate-Limit-Remaining', 'X-Rate-Reset'
]:
assert header in response['headers'],\
"Expected {0} header to be returned".format(header)
def guess_list(self, path='files/names/partial.csv', binary=False):
# guess list method
slist = []
with open(path) as csvfile:
sexreader = csv.reader(csvfile, delimiter=',', quotechar='|')
next(sexreader, None)
i = 0
# string = ""
listnames = list()
for row in sexreader:
name = row[0].title()
name = name.replace('\"','')
listnames.append(name)
# print("len listnames:"+str(len(listnames)))
new = []
for i in range(0, len(listnames), 10): # We must split the list in different lists with size 10
new.append(listnames[i : i+10])
for i in new:
if (self.config['DEFAULT']['genderize'] == 'no'):
jsonlist = Genderize().get(i)
elif (self.config['DEFAULT']['genderize'] == 'yes'):
fichero = open("files/apikeys/genderizepass.txt", "r+")
apikey = fichero.readline().rstrip()
jsonlist = Genderize(user_agent='GenderizeDocs/0.0', api_key=apikey).get(i)
for item in jsonlist:
if ((item['gender'] == None) & binary):
slist.append(2)
elif ((item['gender'] == None) & (not binary)):
slist.append("unknown")
elif ((item['gender'] == "male") & binary):
slist.append(1)
elif ((item['gender'] == "male") & (not binary) ):
slist.append("male")
elif ((item['gender'] == "female") & binary):
slist.append(0)
elif ((item['gender'] == "female") & (not binary) ):
slist.append("female")
return slist
def __init__(self, data, key=None, gender_file=None):
""" Main constructor of the class where the original dataframe
is provided.
:param data: original dataframe
:param key: genderize key (optional)
:param gender_file: file with gender info, used as cache
:type data: pandas.DataFrame
:type key: string
:type gender_file: string (as filepath)
"""
from genderize import Genderize
self.data = data
self.gender = {} # init the name-gender dictionary
self.key = key
self.gender_file = gender_file
# Init the genderize connection
self.connection = Genderize()
if self.key:
self.connection = Genderize(api_key=self.key)
if self.gender_file:
# This file is used as cache for the gender info
# This helps to avoid calling once and again to the API
fd = open(gender_file, "r")
lines = fd.readlines()
fd.close()
# TODO: fix hardcoded code when reading columns and using
# separators
for line in lines:
gender_data = line.split("\t")
self.gender[gender_data[1]] = {
"gender_analyzed_name": gender_data[1],
"gender": gender_data[2]
}
def get_gender_from_given_name(ungendered_imdbs):
names = {}
for imdb in ungendered_imdbs:
if imdb[0] not in names.keys():
names[imdb[0]] = {}
names[imdb[0]]['imdbs'] = []
names[imdb[0]]['imdbs'].append(imdb[1])
#print(names)
#trying to get gender from name file
try:
with open('names_to_gender.csv', 'r', encoding="utf-8") as file:
for line in file:
info = line.split(',')
if info[0] in names:
#print(info[0]+"("+info[1]+") is given name of "+str(names[info[0]]['imdbs']))
for imdb in names[info[0]]['imdbs']:
#print(imdb)
crew[imdb]['gender'] = info[1]
crew[imdb]['gender_probability'] = info[2]
crew[imdb]['source'] = 'Genderize.io'
del names[info[0]]
except:
print('names_to_gender file not found. No problem. Proceeding.')
names_list = list(names.keys())
start = 0
stop = 9
namesfile = open("names_to_gender.csv", 'a', encoding="utf-8")
while (start < len(names_list)):
#print("Getting gender of names "+str(names_list[start:stop+1]))
gendered_names = Genderize().get(names_list[start:stop + 1])
for name in gendered_names:
#print(name['name'])
namesfile.write(
str(name['name']) + ',' + str(name['gender']) + ',' +
str(name['probability']) + ',' + str(name['count']) + '\n')
for imdb in names[name['name']]['imdbs']:
#print(imdb)
crew[imdb]['gender'] = name['gender']
crew[imdb]['gender_probability'] = name['probability']
crew[imdb]['source'] = 'Genderize.io'
start = start + 10
stop = min(stop + 10, len(names) + 1)
time.sleep(5)
namesfile.close()
def find_gender():
f = open('kiev_users.txt', 'r')
users = pickle.load(f)
for user in users:
gender = None
user.full_name = translit(user.full_name, 'ru', reversed=True)
result = Genderize().get(user.full_name.split())
for item in result:
# print user.full_name, item['gender']
# print item['gender']
# if 'probability' in item:
# print item['probability']
if item['gender'] == 'female' or item['gender'] == 'male':
gender = item['gender']
print user.full_name, gender
class GenderizeIO(object):
__name__ = 'genderizeio'
def __init__(self, api_key=None):
self._api = Genderize(api_key=api_key)
def gender(self, artist):
first_name = artist.split(' ')[0]
response = self._api.get([first_name])
if response[0]['gender'] is not None \
and response[0]['probability'] > 0.75 \
and response[0]['count'] > 50:
return response[0]['gender']
else:
return None
def __init__(self, data, key=None, gender_file=None):
""" Main constructor of the class where the original dataframe
is provided.
:param data: original dataframe
:param key: genderize key (optional)
:param gender_file: file with gender info, used as cache
:type data: pandas.DataFrame
:type key: string
:type gender_file: string (as filepath)
"""
from genderize import Genderize
self.data = data
self.gender = {} # init the name-gender dictionary
self.key = key
self.gender_file = gender_file
# Init the genderize connection
self.connection = Genderize()
if self.key:
self.connection = Genderize(api_key=self.key)
if self.gender_file:
# This file is used as cache for the gender info
# This helps to avoid calling once and again to the API
fd = open(gender_file, "r")
lines = fd.readlines()
fd.close()
# TODO: fix hardcoded code when reading columns and using
# separators
for line in lines:
gender_data = line.split("\t")
self.gender[gender_data[1]] = {"gender_analyzed_name": gender_data[1],
"gender": gender_data[2]}
def dumpintoexcel(fullname, salutation, designation, namelist, sociallinks,
countries, imagelinks, last_updated):
try:
from genderize import Genderize
x = Genderize().get([namelist.first])
for y in x:
if y['gender'] != None or y['gender'] != '':
Gender = y['gender']
else:
Gender = 'Unknown'
except:
Gender = 'Unknown'
links = ','.join([str(elem) for elem in sociallinks])
images = ','.join([str(elem) for elem in imagelinks])
if namelist:
first_name = namelist.first
middle_name = namelist.middle
last_name = namelist.last
else:
first_name = ""
middle_name = ""
last_name = ""
infotext = [{
'Country': countries[0],
'Full Name': fullname,
'First Name': first_name,
'Middle Name': middle_name,
'Last Name': last_name,
'Gender': Gender,
'Title': salutation,
'Designation': designation,
'Contact': links,
'Images': images,
'Last Updated': last_updated,
}]
keys = [
'Country', 'Full Name', 'First Name', 'Middle Name', 'Last Name',
'Gender', 'Title', 'Designation', 'Contact', 'Images', 'Last Updated'
]
with open('ScrappedData/PoliticalLeaders.csv', 'a') as dept:
writer = csv.DictWriter(dept, fieldnames=keys)
writer.writerows(infotext)
def genderize_function(name):
#lowercase
name = name.lower()
regex = re.compile('[^a-z ]')
#removal of all non-letters
name = regex.sub('', name)
#names = name.split()
names = name.split()
print names
try:
for x in names:
gen = Genderize(api_key=GENDERIZE_API_KEY).get([x])
print gen
if gen[0]['gender'] != None:
return (gen[0]['gender'],gen[0]['count'],gen[0]['probability'])
return ('unknown','unknown','unknown')
except genderize.GenderizeException:
pass
def get_author_gender():
name = request.args.get('name')
conn = getConnection()
cur = conn.cursor()
query = "select authors.gender, authors.gender_source from authors where authors.name = %s"
cur.execute(query, (name, ))
rows = cur.fetchall()
if not rows:
gender = Genderize(api_key="a619730661a7ce6b4f8e8e6b047046a2").get(
[name.split(" ")[0]])[0]["gender"]
source = "genderize"
insert = "insert into authors (name, gender, gender_source) values (%s, %s, %s)"
cur.execute(insert, (name, gender, source))
else:
gender = rows[0][0]
source = rows[0][1]
print(query, rows)
conn.close()
return {"name": name, "gender": gender, "source": source}
def get_gender(df, column):
name_list = []
genderdict_list = []
gender_list = []
for name in df[column]:
name_list.append(name)
name_list = [name_list[i:i + 10] for i in range(0, len(name_list), 10)]
for element in name_list:
genderdict_list.append(Genderize().get(element))
for gender in genderdict_list:
for key in gender:
gender_list.append(key['gender'])
return gender_list
def get_gender(unique_name):
"""
Request genderize api to get the gender for each name in a list.
If maximum number of requests is reached, return the last data stored.
"""
try:
get_gender = Genderize().get(unique_name)
fp = os.path.join('extern_data', 'gender.json')
f = open(fp, "w")
json.dump(get_gender, f)
f.close()
return get_gender
except GenderizeException:
print("Request limit")
fp = os.path.join('extern_data', 'gender.json')
if os.path.exists(fp):
with open(fp) as data_file:
data_loaded = json.load(data_file)
return data_loaded
def call_genderizeio(auth_df):
unk_df = auth_df[(auth_df['clean_gender'] == 'unknown')
& (auth_df['forename'] != 'NAN') &
(auth_df['forename'].str.len() > 2)]
print('There are ' + str(len(unk_df)) +
' names we dont know about before calling genderize...')
unk_df = pd.DataFrame(unk_df['forename'].drop_duplicates())
r = requests.get("https://api.genderize.io?name=test")
print(r.headers['X-Rate-Limit-Remaining'] +
' genderize calls remaining...!')
if int(r.headers['X-Rate-Limit-Remaining']) >= len(unk_df):
genderize_return = (Genderize().get(unk_df['forename'].to_list()))
unk_df['genderize_return'] = ''
counter = 0
for index, row in unk_df.iterrows():
unk_df.at[index,
'genderize_return'] = genderize_return[counter]['gender']
counter = counter + 1
#call_genderizeio()
unk_df.to_csv('unknown_names.csv')
unk_df = unk_df[(unk_df['genderize_return'] == 'male') |
(unk_df['genderize_return'] == 'female')]
auth_df = pd.merge(auth_df,
unk_df,
how='left',
left_on='forename',
right_on='forename')
auth_df['clean_gender'] = np.where(
auth_df['genderize_return'] == 'male', auth_df['genderize_return'],
auth_df['clean_gender'])
auth_df['clean_gender'] = np.where(
auth_df['genderize_return'] == 'female',
auth_df['genderize_return'], auth_df['clean_gender'])
unk_df = auth_df[(auth_df['clean_gender'] == 'unknown')
& (auth_df['forename'] != 'NAN') &
(auth_df['forename'].str.len() > 2)]
print('After calling the genderize API, there are still ' +
str(len(unk_df)) + ' names which we dont know about!...')
return auth_df
def trouver_sexe_prof(prenom):
sexe = Genderize().get([prenom])
for sex in sexe:
gender = list(sex.values())[1]
#print(gender)
if gender is not None:
probability = list(sex.values())[2]
#print(probability)
if gender == None:
return "impossible à déterminer"
else:
if gender == "female":
try:
if probability >= 0.75:
return "femme"
except IndexError:
return "impossible à déterminer"
elif gender == "male":
try:
if probability >= 0.75:
return "homme"
except IndexError:
return "impossible à déterminer"
def get_genderize(api_key):
""" Returns genderize object, useful when using this script as a module
"""
genderize = Genderize(user_agent='GenderizeDocs/0.0', api_key=api_key)
return genderize
class Gender(Enrich):
""" This class creates three new columns with the gender of
the name provided
"""
def __init__(self, data, key=None, gender_file=None):
""" Main constructor of the class where the original dataframe
is provided.
:param data: original dataframe
:param key: genderize key (optional)
:param gender_file: file with gender info, used as cache
:type data: pandas.DataFrame
:type key: string
:type gender_file: string (as filepath)
"""
from genderize import Genderize
self.data = data
self.gender = {} # init the name-gender dictionary
self.key = key
self.gender_file = gender_file
# Init the genderize connection
self.connection = Genderize()
if self.key:
self.connection = Genderize(api_key=self.key)
if self.gender_file:
# This file is used as cache for the gender info
# This helps to avoid calling once and again to the API
fd = open(gender_file, "r")
lines = fd.readlines()
fd.close()
# TODO: fix hardcoded code when reading columns and using
# separators
for line in lines:
gender_data = line.split("\t")
self.gender[gender_data[1]] = {"gender_analyzed_name": gender_data[1],
"gender": gender_data[2]}
def enrich(self, column):
""" This method calculates thanks to the genderize.io API the gender
of a given name.
This method initially assumes that for the given
string, only the first word is the one containing the name
eg: Daniel Izquierdo <*****@*****.**>, Daniel would be the name.
If the same class instance is used in later gender searches, this stores
in memory a list of names and associated gender and probability. This is
intended to have faster identifications of the gender and less number of
API accesses.
:param column: column where the name is found
:type column: string
:return: original dataframe with four new columns:
* gender: male, female or unknown
* gender_probability: value between 0 and 1
* gender_count: number of names found in the Genderized DB
* gender_analyzed_name: name that was sent to the API for analysis
:rtype: pandas.DataFrame
"""
if column not in self.data.columns:
return self.data
splits = self.data[column].str.split(" ")
splits = splits.str[0]
self.data["gender_analyzed_name"] = splits.fillna("noname")
self.data["gender_probability"] = 0
self.data["gender"] = "Unknown"
self.data["gender_count"] = 0
names = list(self.data["gender_analyzed_name"].unique())
for name in names:
if name in self.gender.keys():
gender_result = self.gender[name]
else:
try:
# TODO: some errors found due to encode utf-8 issues.
# Adding a try-except in the meantime.
gender_result = self.connection.get([name])[0]
except Exception:
continue
# Store info in the list of users
self.gender[name] = gender_result
# Update current dataset
if gender_result["gender"] is None:
gender_result["gender"] = "NotKnown"
self.data.loc[self.data["gender_analyzed_name"] == name, 'gender'] =\
gender_result["gender"]
if "probability" in gender_result.keys():
self.data.loc[self.data["gender_analyzed_name"] == name,
'gender_probability'] = gender_result["probability"]
self.data.loc[self.data["gender_analyzed_name"] == name,
'gender_count'] = gender_result["count"]
self.data.fillna("noname")
return self.data