In [1]:

    

%matplotlib inline


# imports needed for the following examples
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
# import scipy.spatial.distance as distance

# read a local file (path is relative to python's working directory)
# sep, header=True/None
infile = '../data/PO_asof_20150525.txt'
df = pd.read_table(infile, sep="|", thousands=',')

# set column name
df.columns = ['comp_code', 'comp_name', 'vendor_code', 'vendor_name',
              'which_day', 'po', 'amt']

grouped_comp = df.groupby('comp_code').agg({'amt': np.sum,
                                            'po': 'count'})

grouped_comp.rename(columns={'po': 'po_cnt'}, inplace=True)
grouped_comp['po_cnt_log'] = np.log(grouped_comp['po_cnt'])
grouped_comp['amt_log'] = np.log(grouped_comp['amt'])


    

In [2]:

    

fig = plt.figure()
ax = grouped_comp['po_cnt'].plot('box')
# ax.set_xlabel("xx")
ax.set_ylabel('po count')
plt.xticks([1], ['plant'])

fig.savefig('test.png')


    

In [12]:

    

ax = grouped_comp['po_cnt'].plot('hist', alpha=0.5, bins=20)
ax.set_xlabel('po count')
ax.set_ylabel('plant count')


    

    
Out[12]:





<matplotlib.text.Text at 0x10a9382b0>






    

In [11]:

    

ax = grouped_comp['po_cnt_log'].plot('hist', alpha=0.5, bins=30)
ax.set_xlabel('log2(po count)')
ax.set_ylabel('plant count')


    

    
Out[11]:





<matplotlib.text.Text at 0x10a779198>






    

In [62]:

    

np.power(2, 9.363717)
np.log2(894.000000 )


    

    
Out[62]:





9.8041310211833181

In [56]:

    

grouped_comp.describe()


    

    
Out[56]:






  

    

      

      
vendor_cnt
      
amt
      
vendor_cnt_log
      
amt_log
    
  
  

    

      
count
      
82.000000
      
8.200000e+01
      
82.000000
      
82.000000
    
    

      
mean
      
3859.414634
      
9.075405e+08
      
9.363717
      
26.454046
    
    

      
std
      
5588.489159
      
1.484848e+09
      
3.661000
      
4.333323
    
    

      
min
      
1.000000
      
1.030770e+04
      
0.000000
      
13.331435
    
    

      
25%
      
129.500000
      
1.674473e+07
      
7.014306
      
23.996830
    
    

      
50%
      
894.000000
      
1.336340e+08
      
9.803694
      
26.991069
    
    

      
75%
      
5552.250000
      
1.100607e+09
      
12.438853
      
30.017386
    
    

      
max
      
25029.000000
      
6.074226e+09
      
14.611313
      
32.500054
    
  

In [5]:

    

grouped_comp.sort(['po_cnt_log', 'amt_log'], ascending=[False, False])

fig2, ax2 = plt.subplots(nrows=1, ncols=2)
fig2.set_size_inches(24, 8)
plt.subplots_adjust(wspace=0.2)

# 5950|华能海南发电股份有限公司东方电厂
# 4732|华能吉林发电有限公司长春生物质热电厂
# 51A0|华能济南黄台发电有限公司

grouped_comp.plot(kind='scatter', x='po_cnt', y='amt', ax=ax2[0],
                  title='po count vs. amount')
ax2[0].set_xlabel('po count')
ax2[0].set_ylabel('toal amount')

grouped_comp.plot(kind='scatter', x='po_cnt_log', y='amt', ax=ax2[1],
                  title='vendor count vs. log2(amount)')
ax2[1].set_xlabel('log(po count)')
ax2[1].set_ylabel('total amount')


    

    
Out[5]:





<matplotlib.text.Text at 0x107b97e10>






    

In [ ]: