Book a Demo!
CoCalc Logo Icon
StoreFeaturesDocsShareSupportNewsAboutPoliciesSign UpSign In
y33-j3T
GitHub Repository: y33-j3T/Coursera-Deep-Learning
 Path: blob/master/Natural Language Processing with Classification and Vector Spaces/Week 3 - Vector Space Models/utils_vecs.py
14375 views
1
# This is used to tranlate english to french

2


3
import pandas as pd

4
from gensim.models import KeyedVectors

5
import nltk

6
import unicodedata

7
import string

8


9
# Loading in the French embeddings. 

10


11
fr_embeddings = KeyedVectors.load_word2vec_format('wiki.multi.fr.vec')

12
f = open('capitals.txt', 'r').read()

13
set_words = set(nltk.word_tokenize(f))

14


15
def load_translations():

16
    '''

17
    TBD

18
    

19
    '''

20
    dict_fr = pd.read_csv('en-fr.txt', delimiter = ' ')

21
    

22
    en_to_fr = {}

23
    fr_to_vec = {}

24
    for i in range(len(dict_fr)):

25
        en = dict_fr.loc[i][0]

26
        fr = dict_fr.loc[i][1]

27
        if type(en) != float:

28
            en = en.capitalize()

29
        if en in set_words and en not in set(en_to_fr.keys()):

30
            en_to_fr[en] = fr

31
            fr_to_vec[fr] = fr_embeddings[fr]

32
    # Add comments later 

33
    del fr_to_vec['syrienne']

34
    del fr_to_vec['iranienne']

35
    del fr_to_vec['malien']

36
    del fr_to_vec['arménienne']

37
    del fr_to_vec['chilien']

38
    del fr_to_vec['équateur']

39
    en_to_fr['Chile'] = 'chili'

40
    fr_to_vec['chili'] = fr_embeddings['chili']

41
    en_to_fr['Iran'] = 'iran'

42
    fr_to_vec['iran'] = fr_embeddings['iran']

43
    en_to_fr['Turkey'] = 'turquie'

44
    fr_to_vec['turquie'] = fr_embeddings['turquie']

45
    en_to_fr['Syria'] = 'syrie'

46
    fr_to_vec['syrie'] = fr_embeddings['syrie']

47
    en_to_fr['Nigeria'] = 'nigeria'

48
    fr_to_vec['nigeria'] = fr_embeddings['nigeria']

49
    en_to_fr['Mali'] = 'mali'

50
    fr_to_vec['mali'] = fr_embeddings['mali']

51
    fr_to_vec['grece'] = fr_embeddings['grèce']

52
    en_to_fr['Armenia'] = 'arménie'

53
    fr_to_vec['arménie'] = fr_embeddings['arménie']

54
    en_to_fr['Ecuador'] = 'ecuador'

55
    fr_to_vec['ecuador'] = fr_embeddings['ecuador']

56
    en_to_fr['Niger'] = 'niger'

57
    fr_to_vec['niger'] = fr_embeddings['niger']

58
    return en_to_fr, fr_to_vec

59


60
def remove_accents(data):

61
    return ''.join(x for x in unicodedata.normalize('NFKD', data) if x in string.ascii_letters).lower()

62


63