将单列拆分为 Dataframe 中的 4 个不同的单独列

Question

I just need need to split a single column of dataframe to 4 different columns. I tried few steps but didn't worked.

DATA1:

 Dump               
12525 2 153 89-8 Winch
24798 1 147 65-4 Gear
65116 4          Screw 
46456 1          Rowing
46563 5          Nut       

Expected1:

 Item  Qty  Part_no    Description             
12525  2    153 89-8   Winch
24798  1    147 65-4     Gear
65116  4               Screw 
46456  1               Rowing
46563  5               Nut       

DATA2:

 Dump               
12525 2 153 89-8 Winch Gear
24798 1 147 65-4 Gear nuts
65116 X          Screw bolts
46456 1          Rowing rings
46563 X          Nut       

Expected2:

 Item  Qty  Part_no    Description             
12525  2    153 89-8   Winch Gear
24798  1    147 65-4   Gear nuts
65116  X               Screw bolts
46456  1               Rowing rings
46563  X               Nut       

I tried the below code

data_df[['Item','Qty','Part_no','Description']] = data_df["Dump"].str.split(" ", 3, expand=True)

and got the output like 

 Item  Qty  Part_no  Description             
12525  2    153 89-8   Winch
24798  1    147 65-4   Gear
65116  4    Screw 
46456  1    Rowing
46563  5    Nut       

Also I tried with this code but not got the expected output:

data_df[['Item','Qty','Part_no','Description']] = data_df['Dump'].str.extract(r'(\d+)\s+(\S+)\s+(\d*)\s*(.+)$')

Any suggestions, how can i fix this???

Similar to this question: Split the single column to 4 different columns in Dataframe

Answer 1

You could match the data format of the Part_no column in a capture group and make the data in that group optional to keep 4 columns.

(\d+)\s+(\S+)\s+((?:\d+\s+\d+-\d+)?\s*)(.+)$

Regex demo

Example with named capture groups and str.extractall

import pandas as pd

pattern = r'(?m)(?P<Item>\d+)\s+(?P<Qty>\S+)\s+(?P<Part_no>(?:\d+\s+\d+-\d+)?\s*)(?P<Description>.+)$'
items = [("12525 2 153 89-8 Winch Gear\n"
          "24798 1 147 65-4 Gear nuts\n"
          "65116 X          Screw bolts\n"
          "46456 1          Rowing rings\n"
          "46563 X          Nut  ")]

data_df = pd.DataFrame(items, columns=["Dump"])
res = data_df['Dump']\
    .str\
    .extractall(pattern)\
    .fillna('')

print(res)

Output

          Item Qty    Part_no   Description
  match                                    
0 0      12525   2  153 89-8     Winch Gear
  1      24798   1  147 65-4      Gear nuts
  2      65116   X              Screw bolts
  3      46456   1             Rowing rings
  4      46563   X                    Nut