python中pandas数据帧的矩阵表示

Question

i have a data-frame like

 from   to  Amt
 a      b   100
 a      c   200
 a      d   220
 b      a   250
 b      c   300
 b      d   330
 c      a   100
 c      b   120
 c      d   320
 d      a   211
 d      b   980
 d      c   430    

i want to represent it in matrix format like

     a     b     c    d
a    0    100    200  220
b   250    0     300  330
c   100   120    0    320
d   211   980    430   0

How to achieve that..

i have followed Printing Lists as Tabular Data link.But not getting what i was looking for.

Answer 1

You need to pivot your data. Here is an example.

pivot_df = df.pivot(index='from', columns='to', values='Amt')

For doing fractional calculations before hand, you might use groupby() then transform('sum') . It is similar to a SQL window function sum.

df['sums'] =  df.groupby('from')['amt'].transform('sum')
df['frac'] = df['amt'] / df['sums']
df.pivot(index='from', columns='to', values='frac')

Answer 2

You need to pivot the data frame. See http://pandas.pydata.org/pandas-docs/stable/reshaping.html

df.pivot(index="from", columns="to",values="Amt" )

Answer 3

You could also achieve that with pivot_table :

df_pivoted = pd.pivot_table(df, index='from', columns='to', fill_value=0)
print(df_pivoted)

      Amt               
to      a    b    c    d
from                    
a       0  100  200  220
b     250    0  300  330
c     100  120    0  320
d     211  980  430    0