prob = prob + log(0.00004, 2)
# perplex = float(-1)*(float(1)/float(v))*prob
# perplex = 10**perplex
# print perplex
return prob
myset = Set([1, 2, 3, 4, 5])
temp = myset
fp = open("results_unigram", "w+")
Pos_Dict = dict()
Neg_Dict = dict()
for x in myset:
temp.remove(x)
Pos_Dict = get_bigram("pos", "dataset", temp)
# Calculate pos_perplexity
print len(Pos_Dict)
Neg_Dict = get_bigram("neg", "dataset", temp)
print len(Neg_Dict)
# Calculate neg_perplexity
fpath = "dataset/" + str(x)
test_file_p = os.listdir(fpath + "/pos")
test_file_n = os.listdir(fpath + "/neg")
fp.writelines("Test folder:" + str(x) + "\n")
# test positive folder under test folder
y_true = list()
y_pred = list()
Npos = 0
Nneg = 0
fh.close()
'''
myset = Set([1,2,3,4,5]);
temp = myset
#fp = open('results_unigram','w+')
master_dict= dict()
Pos_Dict = dict()
Neg_Dict= dict()
path = 'dataset'
for x in myset:
#temp.remove(x)
Pos_Dict = get_bigram('pos', path, temp)
Neg_Dict = get_bigram('neg', path, temp)
print "1"
master_dict.update(Pos_Dict)
master_dict.update(Neg_Dict)
print "2"
master = ()
master = sorted(master_dict)
print "3"
#print master_dict
master = ()
master = sorted(master_dict)
print "4"
#print master
perplex = float(-1)*(float(1)/float(v))*prob
perplex = 10000*(10**perplex)
#print perplex
return perplex
myset = Set([1,2,3,4,5]);
temp = myset
fp = open('results_bigram','w+')
Pos_Dict_uni= dict()
Pos_Dict_bi= dict()
Neg_Dict_uni= dict()
Neg_Dict_bi= dict()
for x in myset:
temp.remove(x)
Pos_Dict_uni = get_unigram('pos','/home/avj/Documents/NLP/NLP_BinaryClassifier/dataset',temp)
Pos_Dict_bi = get_bigram('pos','/home/avj/Documents/NLP/NLP_BinaryClassifier/dataset',temp)
# Calculate pos_perplexity
print len(Pos_Dict_bi)
print len(Pos_Dict_uni)
Neg_Dict_uni = get_unigram('neg','/home/avj/Documents/NLP/NLP_BinaryClassifier/dataset',temp)
Neg_Dict_bi = get_bigram('neg','/home/avj/Documents/NLP/NLP_BinaryClassifier/dataset',temp)
print len(Neg_Dict_bi)
print len(Neg_Dict_uni)
# Calculate neg_perplexity
fpath = '/home/avj/Documents/NLP/NLP_BinaryClassifier/dataset/'+str(x)
test_file_p = os.listdir(fpath +'/pos')
test_file_n = os.listdir(fpath+'/neg')
fp.writelines("Test folder:"+str(x)+"\n")
#test positive folder under test folder
Pos_Dict_bi = convert(Pos_Dict_uni, Pos_Dict_bi)
