In [1]:

    

import pandas as pd


    

In [2]:

    

data = [{'test': 'vikash|Arpit', 'val': 6},
        {'test': 'vikash_1|arpit|Vinayp', 'val': 3},
        {'test': 'arpit|vinayp', 'val': 2}]


    

In [3]:

    

df = pd.DataFrame.from_dict(data, orient='columns')


    

In [4]:

    

df


    

    
Out[4]:






  

    

      

      
test
      
val
    
  
  

    

      
0
      
vikash|Arpit
      
6
    
    

      
1
      
vikash_1|arpit|Vinayp
      
3
    
    

      
2
      
arpit|vinayp
      
2
    
  

In [5]:

    

chosen_columns = set()
for idx, row in df.iterrows():
    for val in str(row['test']).lower().split('|'):
        chosen_columns.add(val.strip())


    

In [6]:

    

chosen_columns_list = list(chosen_columns)


    

In [7]:

    

chosen_columns_list.sort(key=len, reverse=True)


    

In [8]:

    

chosen_columns_list


    

    
Out[8]:





['vikash_1', 'vinayp', 'vikash', 'arpit']

In [9]:

    

def get_one_hot_encoded_column(col_value):
    col_value = col_value.lower()
    new_col_value = ''
    for val in chosen_columns_list:
        if val in col_value.split('|'):
            col_value = col_value.replace(val, '')
            new_col_value += '1,'
        else:
            new_col_value += '0,'
    return new_col_value[:-1]


    

In [10]:

    

df['test_new'] = df['test'].map(get_one_hot_encoded_column)


    

In [11]:

    

df


    

    
Out[11]:






  

    

      

      
test
      
val
      
test_new
    
  
  

    

      
0
      
vikash|Arpit
      
6
      
0,0,1,1
    
    

      
1
      
vikash_1|arpit|Vinayp
      
3
      
1,1,0,1
    
    

      
2
      
arpit|vinayp
      
2
      
0,1,0,1
    
  

In [12]:

    

df2 = df['test_new'].apply(lambda x: pd.Series([i for i in x.lower().split(',')]))


    

In [15]:

    

df2


    

    
Out[15]:






  

    

      

      
vikash_1
      
vinayp
      
vikash
      
arpit
    
  
  

    

      
0
      
0
      
0
      
1
      
1
    
    

      
1
      
1
      
1
      
0
      
1
    
    

      
2
      
0
      
1
      
0
      
1
    
  

In [16]:

    

df2.columns = chosen_columns_list


    

In [17]:

    

df2


    

    
Out[17]:






  

    

      

      
vikash_1
      
vinayp
      
vikash
      
arpit
    
  
  

    

      
0
      
0
      
0
      
1
      
1
    
    

      
1
      
1
      
1
      
0
      
1
    
    

      
2
      
0
      
1
      
0
      
1
    
  

In [18]:

    

df2.info()


    

    




<class 'pandas.core.frame.DataFrame'>
RangeIndex: 3 entries, 0 to 2
Data columns (total 4 columns):
vikash_1    3 non-null object
vinayp      3 non-null object
vikash      3 non-null object
arpit       3 non-null object
dtypes: object(4)
memory usage: 176.0+ bytes

In [19]:

    

df2 = df2.apply(pd.to_numeric)


    

In [20]:

    

df2.info()


    

    




<class 'pandas.core.frame.DataFrame'>
RangeIndex: 3 entries, 0 to 2
Data columns (total 4 columns):
vikash_1    3 non-null int64
vinayp      3 non-null int64
vikash      3 non-null int64
arpit       3 non-null int64
dtypes: int64(4)
memory usage: 176.0 bytes

In [28]:

    

df_new = pd.concat([df, df2], axis=1)


    

In [29]:

    

df_new.drop(['test', 'test_new'], inplace=True, axis=1)


    

In [30]:

    

df_new


    

    
Out[30]:






  

    

      

      
val
      
vikash_1
      
vinayp
      
vikash
      
arpit
    
  
  

    

      
0
      
6
      
0
      
0
      
1
      
1
    
    

      
1
      
3
      
1
      
1
      
0
      
1
    
    

      
2
      
2
      
0
      
1
      
0
      
1
    
  

In [ ]: