data_pre-processing.py

#=====================================================================================
#this script will eliminate the unwanted portions of feature file-
#generated by OpenSmile and store them in Numpy array formatted text files.

# Written by Mahbub Ul Alam
#=====================================================================================

#!/usr/bin/python
import os 
import sys

#location of the feature file generated from openSmile.
#if you are using newer version of python then please use "input" instead of "raw_input" .
file_path = raw_input("Enter OpenSmile generated Feature containing CSV File location: ")

#if K-fold cross-validation is needed to be performed on data
#then please select All (A).
data_type = raw_input("Enter output Data Type, T for Training, D for Development, A for All: ")


# processed data will be saved in text files.
if(data_type=="T"):
    output_file_name  = open("training_data.txt",'w')
    print("Data will be generated shortly, please check training_data.txt")
elif(data_type=="D"):
    output_file_name  = open("development_data.txt",'w')
    print("Data will be generated shortly, please check development_data.txt file")
else:
    output_file_name  = open("processed_data.txt",'w')
    print("Data will be generated shortly, please check processed_data.txt file")
    
sys.stdout = output_file_name

#data will be saved in Numpy array format
input_file = open(file_path)
count=-1
token=-1
for line in input_file.read().split('\n'):
    count=-1
    token=-1
    if(line!='') :
        if(line[0]!='@'):
            for value in line.split(","):
                count+=1
                if(count==0):
                    for word in value.split("_"):
                        token+=1
                        if(token==0):
                            break
                break;
            line = line.replace(value,word)
            line = line.replace('\'','')
            line = line.replace(",?",'')
            print (line)