2

I want to construct a Term-Document matrix in python for Arabic language, I used CountVectorizer(), but it gives the documents in the column, and terms in the row. I want the terms to be in columns and documents in rows, I tried to transpose the matrix, but it gives wrong values, here is the code that i have used.

corpus = [ 'القدرة على التفكير و الحل', 'القدرة على التعلم و القدرة على التفكير في المسائل', 'القدرة على التعلم في حل المسائل']
vectorizer = CountVectorizer()
X = vectorizer.fit_transform(corpus)
v = X.toarray()
Mohammed ALDossari
  • 47
  • 1
  • "I tried to transpose the matrix, but it gives wrong values" What is wrong about them? – juanpa.arrivillaga Feb 26 '19 at 16:53
  • It gives wrong values, is there a way to just create the terms in rows and documents in columns ? – Mohammed ALDossari Feb 26 '19 at 17:01
  • **What are the wrong values?** What values did you expect? The `CountVectorizer` assumes each object in your corpus list is a document. It will always give you documents as rows and terms in columns. The transpose will give you the transpose. What about the transpose is not working for you? What, **exactly** is the output you require? – juanpa.arrivillaga Feb 26 '19 at 17:03

0 Answers0