Big-Data---Sentiment-Analysis/classifier_test.py at main · michellevarghese/Big-Data---Sentiment-Analysis · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
from pyspark import SparkContext
from pyspark.streaming import StreamingContext
from pyspark.sql.session import SparkSession

from sklearn.metrics import precision_score
from sklearn.metrics import recall_score
from sklearn.metrics import f1_score

from pyspark.ml.feature import HashingTF, IDF, Tokenizer
from pyspark.ml.feature import StringIndexer
from pyspark.ml import Pipeline
from pyspark.ml.classification import LogisticRegression
from pyspark.ml.evaluation import BinaryClassificationEvaluator

from sklearn.feature_extraction.text import HashingVectorizer
from sklearn.linear_model import SGDClassifier
from sklearn.neural_network import MLPClassifier
from sklearn.metrics import confusion_matrix

from sklearn.metrics import accuracy_score
from sklearn.model_selection import train_test_split

from pyspark.sql.functions import col
from pyspark.sql.types import StringType,BooleanType,DateType

import numpy as np
import pandas as pd
import json
import pickle

count=0
filename1 = "mlb_model.pkl"
filename2 = "sgd_model.pkl"
filename3 = "perceptron_model.pkl"
filename4 = "mlp_model.pkl"
cm_file = "confusion_matrix.txt"
scores_mlb = "scores_mlb.txt"
scores_sgd = "scores_sgd.txt"
scores_perceptron = "scores_perceptron.txt"
scores_mlp = "scores_mlp.txt"


def preprocess(batch):
	for i in list(batch.collect()):
		if i == []:
			continue
		else:
			global count
			count+=1

			fopen1 = open(filename1,"rb")
			mlb = pickle.load(fopen1)

			fopen2 = open(filename2,"rb")
			sgd = pickle.load(fopen2)

			fopen3 = open(filename3,"rb")
			perceptron = pickle.load(fopen3)

			fopen4 = open(filename4,"rb")
			mlp = pickle.load(fopen4)

			fopen5 = open(cm_file,"a+")
			fopen6 = open(scores_mlb,"a+")
			fopen7 = open(scores_sgd,"a+")
			fopen8 = open(scores_perceptron,"a+")
			fopen9 = open(scores_mlp,"a+")


			test_set = json.loads(i)
			test_set = spark.createDataFrame(test_set.values())

			tokenizer = Tokenizer(inputCol="feature1", outputCol="words")
			hashtf = HashingTF(numFeatures=2**7, inputCol="words", outputCol='tf')
			idf = IDF(inputCol='tf', outputCol="features", minDocFreq=5)
			label_stringIdx = StringIndexer(inputCol = "feature0", outputCol = "label")

			pipeline = Pipeline(stages=[tokenizer, hashtf, idf, label_stringIdx])
			pipelineFit = pipeline.fit(test_set)
			test_df = pipelineFit.transform(test_set)

			a = np.array(test_df.select("features").collect()).reshape(-1, 128)
			y = np.array(test_df.select("feature0").collect()).reshape(-1)


			print("Batch : ",count)

			predictions3 = mlb.predict(a)
			prednp3=np.array(predictions3)
			np.savetxt(cm_file,confusion_matrix(y,prednp3))
			fopen6.write(str(precision_score(y, prednp3,average='weighted'))+ " ")
			fopen6.write(str(recall_score(y, prednp3,average='weighted')) + " ")
			fopen6.write(str(f1_score(y, prednp3,average='weighted')) + "\n")


			predictions1 = sgd.predict(a)
			prednp1=np.array(predictions1)
			np.savetxt(cm_file,confusion_matrix(y,prednp1))
			fopen7.write(str(precision_score(y, prednp1,average='weighted'))+ " ")
			fopen7.write(str(recall_score(y, prednp1,average='weighted') )+" ")
			fopen7.write(str(f1_score(y, prednp1,average='weighted') )+ "\n")

			predictions2 = perceptron.predict(a)
			prednp2 = np.array(predictions2)
			np.savetxt(cm_file,confusion_matrix(y,prednp2))
			fopen8.write(str(precision_score(y, prednp2,average='weighted')) + " ")
			fopen8.write(str(recall_score(y, prednp2,average='weighted')) + " ")
			fopen8.write(str(f1_score(y, prednp2,average='weighted')) + "\n")


			predictions4 = mlp.predict(a)
			prednp4 = np.array(predictions4)
			np.savetxt(cm_file,confusion_matrix(y,prednp4))
			fopen9.write(str(precision_score(y, prednp4,average='weighted')) + " ")
			fopen9.write(str(recall_score(y, prednp4,average='weighted')) + " ")
			fopen9.write(str(f1_score(y, prednp4,average='weighted')) + "\n")

			fopen1.close()
			fopen2.close()
			fopen3.close()
			fopen4.close()
			fopen5.close()
			fopen6.close()
			fopen7.close()
			fopen8.close()
			fopen9.close()


# Create a local StreamingContext with two working thread and batch interval of 1 second
sc = SparkContext()
spark=SparkSession(sc)
ssc = StreamingContext(sc, 1)
sc.setLogLevel("OFF")

lines = ssc.socketTextStream("localhost", 6100)

words = lines.flatMap(lambda line: line.split('\n'))
words.foreachRDD(preprocess)

ssc.start()             # Start the computation
ssc.awaitTermination()