Python: data structure (6)

Abstract: Storage of 2-dimension data into nested dictionary, numpy matrix, and pandas data frame.

The room used:

Size of dictionary (1000x1000) is 48 kB

Size of numpy matrix ((1000, 1000)) is 80 byte

Size of pandas data frame ((1000, 1000)) is 64 byte

Here is the codes:

# -*- coding: utf-8 -*-

"""

Created on Thu Mar 31 10:20:30 2016

@author: yuan

"""

import numpy as np

import pandas as pd

import random

import sys

#get a nested dictionary

def nested_dict(Dimension=100):

data={}

for x in xrange(Dimension):

rowname='A'+str(x)

data[rowname]={}

for y in xrange(Dimension):

colname='B'+str(y)

data[rowname][colname]=random.randint(0,1000)

return data

#get a numpy matrix

def numpy_matrix(Dimension=100):

data=[]

for x in xrange(Dimension):

a=[]

for y in xrange(Dimension):

a.append(random.randint(0,100))

data.append(a )

data=np.array(data)

#print data

return data

#get panda data frame

def pandas_dataframe(Dimension=100):

data=pd.DataFrame(nested_dict(Dimension))

data=data.transpose()

return data

#

d=1000

dict_matrix=nested_dict(d)

size=int(sys.getsizeof(dict_matrix))/1024

print "Size of dictionary (%sx%s) is %d kB" % (d, d,size)

np_matrix=numpy_matrix(d)

#print np_matrix

size=int(sys.getsizeof(np_matrix))

print "Size of numpy matrix (%s) is %d byte" % (np_matrix.shape,size)

df_matrix=pandas_dataframe(d)

#print df_matrix

size=int(sys.getsizeof(df_matrix))

print "Size of pandas data frame (%s) is %d byte" % (df_matrix.shape,size)

print 'ok'