在这两种形式之间转换有好的现有方法吗？

+--------+-------------+
| FRUIT  | ATTRIBUTES  |
+--------+-------------+
| banana | long|yellow |
+--------+-------------+
| kiwi   | brown|oval  |
+--------+-------------+

和

+--------+-----------+
| FRUIT  | ATTRIBUTE |
+--------+-----------+
| banana | long      |
+--------+-----------+
| banana | yellow    |
+--------+-----------+
| kiwi   | brown     |
+--------+-----------+
| kiwi   | oval      |
+--------+-----------+

我目前正在迭代解包过程的行，我听说不鼓励这样做。

import pandas as pd

packed = pd.DataFrame([['banana', 'long|yellow'], 
                       ['kiwi', 'brown|oval']], 
                      columns=['FRUIT', 'ATTRIBUTES'])

pack_delim = '|'

per_fruit_frames = []

for row in packed.itertuples(index=True, name='Pandas'):
    row_attribs = row.ATTRIBUTES
    row_attribs_split = row_attribs.split(pack_delim)
    row_attribs_series = pd.Series(row_attribs_split)
    ras_len = len(row_attribs_split)
    fruit_rep = [row[1]] * ras_len
    frs = pd.Series(fruit_rep)
    temp = pd.concat([frs, row_attribs_series], axis=1)
    per_fruit_frames.append(temp)

unpacked = pd.concat(per_fruit_frames)

unpacked.columns = packed.columns