prediction_pipeline.py

import numpy as np
import pandas as pd
import re
import string
import pickle


from nltk.stem import PorterStemmer
ps = PorterStemmer()


# Load model
with open('static/model/model.pickle', 'rb') as f:
    model = pickle.load(f)


# Load stopwords
with open('static/model/corpora/stopwords/english', 'r') as file:
    sw = file.read().splitlines()


# Load vocabulary
    vocab = pd.read_csv('static/model/vocabulary.txt', header = None)
tokens = vocab[0].tolist()


def remove_punctuations(text):
    for punctuation in string.punctuation: 
        text = text.replace(punctuation, '')
    return text


def preprocessing(text):
    data = pd.DataFrame([text], columns = ['tweet'])
    data["tweet"] = data["tweet"].apply(lambda x: " ".join(x.lower() for x in x.split()))
    data["tweet"] = data['tweet'].apply(lambda x: " ".join(re.sub(r'^https?:\/\/.*[\r\n]*', '', x, flags=re.MULTILINE) for x in x.split()))
    data["tweet"] = data["tweet"].apply(remove_punctuations)
    data["tweet"] = data['tweet'].str.replace('\d+', '',regex=True)
    data["tweet"] = data["tweet"].apply(lambda x: " ".join(x for x in x.split() if x not in sw))
    data["tweet"] = data["tweet"].apply(lambda x: " ".join(ps.stem(x) for x in x.split()))
    return data['tweet']


def vectorizer(ds):
    vectorized_lst = []
    for sentence in ds:
        sentence_lst = np.zeros(len(tokens))
        for i in range(len(tokens)):
            if tokens[i] in sentence.split():
                sentence_lst[i] = 1
        vectorized_lst.append(sentence_lst) 
    vectorized_lst_new = np.asarray(vectorized_lst, dtype=np.float32)   
    return vectorized_lst_new


def get_prediction(vectorized_text):
    prediction = model.predict(vectorized_text)
    if prediction == 1:
        return 'negative'
    else:
        return 'positive'