Python FastText: How to create a corpus from a Dataframe Column

Question

• I need to create a corpus for my Email Classifer . Right now am Using fasttext 0.8.3 but it expects text file as a input whereas i need to pass a dataframe as an input .
  

• It shows error while i am Using following Code :-

  ```

  import fasttext

  x_val = df['Message']  
  y_val = df['Categories']  
  model = fasttext.skipgram(x_val, y_val)  
  print model.words
  
  TypeError:
  <ipython-input-105-58241a9688b5> 
   <module>() 
  ----> 1 model = fasttext.skipgram(x_val, y_val) 
        2 print model.words # list of words in dictionary 
        fasttext/fasttext.pyx in fasttext.fasttext.skipgram (fasttext/fasttext.cpp:6451)() 
        fasttext/fasttext.pyx in fasttext.fasttext.train_wrapper (fasttext/fasttext.cpp:5223)() 
       /root/anaconda2/lib/python2.7/genericpath.pyc in isfile(path) 
             35 """Test whether a path is a regular file""" 
             36 try: 
        ---> 37 st = os.stat(path) 
             38 except os.error: 
             39 return False 
       TypeError: coercing to Unicode: need string or buffer, Series found
  
  ```
  

  • In above code the df['Message'] and df['Categories'] are the dataframe column in which it contains mails and the category respectively .
  • There are 30123 mails in the dataframe .
  • I already go through the Documentation of fasttext but i dont find someting useful.
    

  Fasttext Tutorial refrence

  Thanks for the Help.