import csv # using Python module

f = open('example_1.csv','rb') # use binary mode if on MS windows
d = [i for i in csv.reader(f) ] # use list comprehension to read from file
f.close() # close file
print d

d[0].append('Last Name')
d[0].append('First Name')
print d[0]

for row in d[1:]: # start at 1st, not 0th column. Each row is a list
    first,last= row[0].split() # split on white-space
    row.append(last) # append to each row
    row.append(first)
    
print d

#%qtconsole

f = open('example_1_out.csv','wb') # write mode binary
fw = csv.writer(f) # create csv writer
fw.writerows(d) 
f.close() # close file 

import numpy as np 

d = np.loadtxt('example_1.csv',delimiter=',',dtype=str)

print d
print d.dtype

dt = [('name',   'S64'),   # The first element of the tuple is our name for each respective column
      ('dob',    'S64'),   # and the second element is the numpy dtype we want for that column.
      ('years',  'int'),   # Here we want years as an integer, not a string
      ('degree', 'S64'), ]

d = np.loadtxt('example_1.csv',delimiter=',',dtype=dt,skiprows=1) # skip the header row
print d

print d['years'].mean() # using numpy arrays

import string

n=map(string.split,d['name'])
w=array([tuple(i)+tuple(j) for i,j in zip(d,n)],    # list comprehension glues tuple-ized rows together
         dtype=dt+[('first','S64'),('last','S64')]) # append new dtypes to existing list of dtypes

# the comments are set to '' to avoid hash marks on the first line.
np.savetxt('np_output.csv',w,delimiter=',',fmt='%s',header='name,dob,years,degree,first,last',comments='')

import pandas as pd

d = pd.read_csv('example_1.csv')
print d
print type(d)

print d.columns

print d.DOB # this works great when the column header name has no spaces in it.
print d['Name'] # you can also refer to columns using this syntax

d['name']=d['Name '] # easily create extra column
print d.name         # now you can access this column using this syntax

d = pd.read_csv('example_1.csv',dtype={'Name ':'S64','DOB':'S64','Years':int,'Degree':'S64'})
print d

print d.Years.mean()

d = pd.read_csv('example_1.csv',dtype={'Name':'S64',
                                       'DOB':'S64',
                                       'Years':int,
                                       'Degree':'S64'},parse_dates=[1])

# difference in birthdays between Alice Jones and John Book
print d.DOB[0] -  d.DOB[2]

%qtconsole

d['first']=map(lambda x:string.split(x)[0],d['Name'])
d['last']=map(lambda x:string.split(x)[1],d['Name'])
print d

print d

import pandas.io.sql as pd_sql
import sqlite3 as sql # sqlite3 is built into Python

con = sql.connect("example_1.db")
pd_sql.write_frame(d,'data',con) # write to DB as table named "data"
con.close()