如何在给定的数据帧中划分列“位置”？

Question

I am working on a dataset where a columns is named as header. The value are as mentioned.

df = pd.DataFrame(data={"location":["düsseldorf, nordrhein-westfalen, germany",
                                    "durbanville , cape town, cape town , south africa"]})

I want to divide this column in ['city', 'state', 'country'] . Note the second row has duplicates.

I have tried the below but this does not deal with duplicates:

location = df.location.str.split(', ', n=2, expand=True)

location.columns = ['city', 'state', 'country']

Answer 1

You can use the unique_everseen recipe available in the itertools docs , which is also available in 3rd party libraries such as toolz.unique .

The logic can be incorporated in a list comprehension which iterates df['location'] . This is likely to be more efficient than Pandas string-based methods, which do not offer vectorised functionality.

from toolz import unique

res = pd.DataFrame([list(unique(map(str.strip, i.split(',')))) for i in df['location']])

res.columns = ['city', 'state', 'country']

print(res)

          city                state       country
0   düsseldorf  nordrhein-westfalen       germany
1  durbanville            cape town  south africa

Answer 2

You can limit yourself only with pandas to handle this problem:

import pandas as pd

pd.set_option('display.max_columns', None)
pd.set_option('display.max_rows', None) 

data_all=(['düsseldorf', 'nordrhein-westfalen', 'germany', 'durbanville', 'cape town', 'south africa'])
dfe = [[], [], []]

i = 0
j = 1
k = 2

while i < len(data_all):
  dfe[0].append(data_all[i])
  i += 3
while j < len(data_all):
  dfe[1].append(data_all[j])
  j += 3
while k < len(data_all):
  dfe[2].append(data_all[k])
  k += 3

d = {'city': dfe[0], 'state': dfe[1], 'country': dfe[2]}
df = pd.DataFrame(data=d)
print(df)

Result:

          city                state       country
0   düsseldorf  nordrhein-westfalen       germany
1  durbanville            cape town  south africa

But actually I did not understand why you want to use duplicates, if you only have 3 columns: city, state and country.