def main():
dict = {};
idf = {};
st = "";
ntdocs = 1.0;
docs = os.listdir('papers');
ndocs = len(docs);
n = 0
for i in range(ndocs):
n = n + 1
print docs[i]
print n
f = 'papers' + '/' + docs[i];
statinfo = os.stat(f);
if statinfo.st_size <= 152576:
fp = open(f,"r");
st = fp.read();
dict = tfCalculator.findtf(st);
s = docs[i].strip('.txt');
s = 'pre/' + s + '.json'
f = open(s,"w");
f.write(json.dumps(dict));
f.close();
dict = {};
fp.close();
def main():
#Enter the path of the directory
ntf = tl1 = [];
dict = idf = {};
st = "";
ntdocs = 1;
docs = os.listdir("C:\\Python27\\sampleset");
ndocs = len(docs);
for i in range(ndocs):
f = 'C:\\Python27\\sampleset\\' + docs[i];
fp = open(f,"r");
st = fp.read();
dict = tfCalculator.findtf(st);
#Normalising term frequency
m = max(dict.values()) + 0.0;
for j,v in dict.items():
dict[j] = v/m;
ntf.append(dict); #ntf = [{'word':ntf}]
fp.close();
#Calculating Inverse Document Frequency idf
for i in range(ndocs):
tl1 = ntf[i].keys();
for j in range(len(tl1)):
if idf.has_key(tl1[j]) == False:
for k in range(i+1,ndocs):
if ntf[k].has_key(tl1[j]) == True:
ntdocs = ntdocs + 1;
idf[tl1[j]] = math.log10(1.0 * (ndocs / ntdocs));
ntdocs = 1;
#Input File for Bayon Clustering Tool
fp = open("input.txt","w");
for i in range(ndocs):
fp.write(docs[i]);
fp.write('\t');
tl1 = ntf[i].keys();
for j in range(len(tl1)):
fp.write(tl1[j]);
fp.write('\t');
tfidf = ntf[i].get(tl1[j]) * idf.get(tl1[j]);
st = str(tfidf);
fp.write(st);
fp.write('\t');
fp.write('\n');
fp.close();
