Python和Pandas：匹配和转置

Question

The original table is like:

ID Name Sex Weight Date
1  Bill M   68.1   2015-01-01
2  Mary F   45.2   2015-01-01
3  Gary M   89.2   2015-01-01
1  Bill M   68.3   2015-01-02
2  Mary F   45.3   2015-01-02
3  Gary M   89.1   2015-01-02

---

The desired output is:

ID Name Sex 2015-01-01 2015-01-02
1  Bill M   68.1   68.3
2  Mary F   45.2   45.3
3  Gary M   89.2   89.1

---

Is there any quick way to get the desired data frame? I tried to create a new data frame and append row by row by using ret.loc[i] = row_list, but it's too slow when the data set gets big.

Answer 1

You can use pivot_table :

print df.pivot_table(index=['ID', 'Name','Sex'], columns='Date', values='Weight')
Date         2015-01-01  2015-01-02
ID Name Sex                        
1  Bill M          68.1        68.3
2  Mary F          45.2        45.3
3  Gary M          89.2        89.1

And if you need reset_index use strftime :

df = df.pivot_table(index=['ID', 'Name','Sex'], columns='Date', values='Weight')
df.columns = df.columns.strftime('%Y-%m-%d')

#if you need sorting columns
df= df.sort_index(axis = 1)
df = df.reset_index()

print df
   ID  Name Sex  2015-01-01  2015-01-02
0   1  Bill   M        68.1        68.3
1   2  Mary   F        45.2        45.3
2   3  Gary   M        89.2        89.1

Or:

print df.pivot_table(index=['ID', 'Name','Sex'], columns='Date', values='Weight').reset_index()
Date  ID  Name Sex  2015-01-01 00:00:00  2015-01-02 00:00:00
0      1  Bill   M                 68.1                 68.3
1      2  Mary   F                 45.2                 45.3
2      3  Gary   M                 89.2                 89.1