在一栏熊猫中用最少的字符求值

Question

在df‘the’列中，我希望找到每行字符数最少的单词，并将该值放在df‘’lab‘中。

电流输出

group           stem                                      lab
-----------------  --------- ---------------------  -------------
notif           notify, notified                                nan
face-to-fac     face-to-face                                    nan
propo           proposed                                        nan
lift            lifted                                          nan
govern          governed, governing, government, governance     nan
constitut          constitute, constitutional                   nan
exp                expedient                                    nan
prev               preventing, prevent                          nan
tre                treat, treatment                             nan
work               working, works                               nan

预期产出

group           stem                                      lab
-----------------  --------- ---------------------  -------------
notif           notify, notified                                notify
face-to-fac     face-to-face                                    face-to-face
propo           proposed                                        proposed
lift            lifted                                          lifted
govern          governed, governing, government, governance     governed
constitut       constitute, constitutional                      constitute
exp             expedient                                       expedient
prev            preventing, prevent                             prevent
tre             treat, treatment                                treat
work            working, works                                  works 

试过

df['lab'] = df.stem.str.split(',').apply(lambda x: min(x, key=len))

但它的付出

TypeError: 'float' object is not iterable

Answer 1

您可以使用str.split从stem和apply中的字符串中获取字符串列表，并将内置的min函数设置为key=len，以便返回最短的字符串：

df['lab'] = df.stem.str.split(',').apply(lambda x: min(x, key=len))

某些行的结果：

        group                                            stem         lab
0       notif                                notify, notified      notify
1  ace-to-fac                                      ace-to-fac  ace-to-fac
2      govern     governed, governing, government, governance    governed