def kmeans_optimal(k, news):
matrix = tfidf(news)
centroid = start_optimal(k, matrix)
ii = 0
while ii < 10:
cluster = kernel_cos(centroid, matrix, k)
print rss(cluster, centroid)
centroid = new(cluster)
ii += 1
if ii == 10:
output(cluster, matrix, news)
def kmeans_euc(k, news):
matrix = tfidf(news)
centroid = start(k, matrix)
ii = 0
while ii < 10:
cluster = kernel_euc(centroid, matrix, k)
print rss(cluster, centroid)
centroid = new(cluster)
ii += 1
if ii == 10:
out = token__.labelize(news)
output(cluster, matrix, out)
def m_hot_words_tfidf(data):
printv('Calculateing idf score.')
idf = tf_idf.idf([v[ID_M_BOW] for v in data.values()])
all_result = dict()
for id, bow in data.items():
printv('Calculateing tf-idf score for ' + id)
result = list()
for term in set(bow[ID_M_BOW]):
tf = tf_idf.tf(term, bow[ID_M_BOW])
result.append((term, round(tf_idf.tfidf(term, tf, idf), ROUND_DIGITS)))
all_result[id] = dict()
for word, score in sorted(result, key=lambda x: x[1], reverse=True)[:RESULT_LENGTH]:
all_result[id][word] = score
return all_result
def data_x():
# data=pd.read_csv('quora_duplicate_questions.tsv',delimiter='\t',encoding='utf-8')
# data=data.iloc[2000:2500,:]
str1 = input('Enter your ques 1 - ')
str2 = input('enter your ques 2 - ')
daa = [[0, 1, 2, str1, str2]]
data = pd.DataFrame(daa)
x = data.iloc[:, 1:5]
# y=data.iloc[:,5].values
x = rmpun(x)
x = lower(x)
dataf = pd.DataFrame()
temp = pd.DataFrame()
dif,diff_words,start_word,common_word,ratio,sort_ratio,jcc_sim,wo_sh=[],[],[],[],[],[],[],[]
for i in range(len(x)):
bf.diff_len(str(x.iloc[i, 2]), str(x.iloc[i, 3]))
bf.diff_words(str(x.iloc[i, 2]), str(x.iloc[i, 3]))
bf.start_words(str(x.iloc[i, 2]), str(x.iloc[i, 3]))
bf.common_words(str(x.iloc[i, 2]), str(x.iloc[i, 3]))
af.fuzz_ratio(str(x.iloc[i, 2]), str(x.iloc[i, 3]))
af.fuzz_sort(str(x.iloc[i, 2]), str(x.iloc[i, 3]))
af.jaccard(str(x.iloc[i, 2]), str(x.iloc[i, 3]))
af.word_share(str(x.iloc[i, 2]), str(x.iloc[i, 3]))
pos_data = af.parts_of_speech(str(x.iloc[i, 2]), str(x.iloc[i, 3]))
temp = temp.append(pos_data, ignore_index=True, sort=False)
dataframe = tf_idf.tfidf(str(x.iloc[i, 2]), str(x.iloc[i, 3]))
print(i)
dif, diwords, start_word, common_word = bf.all_data()
ratio, sort_ratio, jcc_sim, wo_sh = af.all_data()
dataf['deff_len'] = dif
dataf['diff_words'] = diwords
dataf['start_words'] = start_word
dataf['common'] = common_word
dataf['fuzz_ratio'] = ratio
dataf['fuzz_sort'] = sort_ratio
dataf['juccard'] = jcc_sim
dataf['word_share'] = wo_sh
dataf = pd.concat([dataf, temp], axis=1)
dataf = pd.concat([dataf, dataframe], axis=1)
# y=np.array(y)
# dataf['target']=y
dataf.to_csv('aaa.csv', index=False)
return dataf
#from CNN_improved import ConvNet
import os
from tf_idf import tfidf
def loadCorpus(folder):
corpus = []
fileList = []
for subdir, dirs, files in os.walk(folder):
for file in files:
path = os.path.join(subdir, file)
if path[-3:] == 'txt':
ticker = path[len(folder) + 1:]
ticker = ticker[:ticker.index('/')]
year = path[path.index('_') + 1:path.index('_') + 5]
fileList.append(ticker + '-' + year)
report = open(path, 'r').read()
corpus.append(report)
return corpus, fileList
corpus, fileList = loadCorpus('data-part1')
data = tfidf(corpus)
data.to_csv('tfidf.csv')
with open("documentNames.txt", "w") as f:
for fileName in fileList:
f.write(str(fileName) + "\n")
#todo modified this if statement
#print("query we are looking at: " + str(query))
for compword in compwords:
if compword in query and parameters.use_booleanSearch:
booleanSearch.constructList(collection, query)
parameters.use_blindRelevance = False
ranBooleanResults = True
# create accumulators and other data structures
accum = {}
filenames = []
tfidfterms = {}
p = porter.PorterStemmer()
sw = StopWord.StopWord()
t = thesaurus.Thesaurus()
tfidf = tf_idf.tfidf()
# get N
f = open(collection + "_index_N", "r")
N = eval(f.read())
f.close()
# get document lengths/titles
titles = {}
f = open(collection + "_index_len", "r")
lengths = f.readlines() #an array of all the file titles and their lengths
f.close()
titleScore = 0
# get index for each term and calculate similarities using accumulators
def rocchio(k, news):
matrix = tfidf(news)
centroid = start_optimal(k, matrix)
kernel(centroid, matrix, news)
