• R/O
  • SSH

Commit

Tags
No Tags

Frequently used words (click to add to your profile)

javac++androidlinuxc#windowsobjective-ccocoa誰得qtpythonphprubygameguibathyscaphec計画中(planning stage)翻訳omegatframeworktwitterdomtestvb.netdirectxゲームエンジンbtronarduinopreviewer

Commit MetaInfo

Revisão2c007f732b7bab7aa98c765d88647b0014c2bdcf (tree)
Hora2015-03-26 01:36:49
AutorLorenzo Isella <lorenzo.isella@gmai...>
CommiterLorenzo Isella

Mensagem de Log

A simple script to convert the test and train datasets (without the target values!) to a numerical matrix based on the term frequency–inverse document frequency.

Mudança Sumário

Diff

diff -r 7eded1ec4fda -r 2c007f732b7b Python-codes/convert-to-tf-idf.py
--- /dev/null Thu Jan 01 00:00:00 1970 +0000
+++ b/Python-codes/convert-to-tf-idf.py Wed Mar 25 17:36:49 2015 +0100
@@ -0,0 +1,43 @@
1+#! /usr/bin/env python
2+
3+
4+
5+import pandas as pd
6+import numpy as np
7+from sklearn import ensemble, feature_extraction, preprocessing
8+
9+
10+# import data
11+train = pd.read_csv('train.csv')
12+test = pd.read_csv('test.csv')
13+#sample = pd.read_csv('sampleSubmission.csv')
14+
15+# drop ids and get labels
16+labels = train.target.values
17+#labels2=np.copy(labels)
18+train = train.drop('id', axis=1)
19+train = train.drop('target', axis=1)
20+test = test.drop('id', axis=1)
21+
22+# transform counts to TFIDF features
23+tfidf = feature_extraction.text.TfidfTransformer()
24+train = tfidf.fit_transform(train).toarray()
25+test = tfidf.transform(test).toarray()
26+
27+#labels=labels.reshape(-1,1)
28+
29+# train=np.hstack((train,labels))
30+
31+# train=pd.DataFrame(train)
32+# test=pd.DataFrame(test)
33+
34+
35+np.savetxt("train-tfidf.csv", train, delimiter=",")
36+np.savetxt("test-tfidf.csv", test, delimiter=",")
37+
38+# train.to_csv("train-tfidf.csv", train)
39+# test.to_csv("test-tfidf.csv", test)
40+
41+
42+
43+print "So far so good"