w2v-average.py

from gensim.models import Word2Vec
import re

def tokenize(sentence):
    remove_dots = re.sub("[.]", "", sentence.lower())
    return re.findall("[A-Za-z]{2,}", remove_dots)

def w2v_average(sentence, w2v_model):
    words = tokenize(sentence.lower())
    sum_w2v = 0
    count_w2v = 0
    for word in words:
        if word in w2v_model.wv.vocab: 
            sum_w2v += w2v_model.wv[word] 
            count_w2v += 1
    if count_w2v:
        return sum_w2v/count_w2v
loaded_w2v = Word2Vec.load('word2vec.bin')
print(w2v_average('VirginAmerica SFO PDX schedule is still MIA ', loaded_w2v))