adding multiple rows in df based on condition

Question

in my df i want to add row when its meets specific condition,df has thousands of rows.In id column when a new value of id is started i want to add new row copying same id value. df is as as below

id                        name                              value
0000                    length                                46
0000                    status                              completed
0000                    segment                              21
1111                     tp                                 0.1
1111                     x                                  56
2222                     point                              23.01
2222                     x                                  50
2222                     y                                  40

expected output is

     id                        name                              value
    0000                       type                                description #new row
    0000                    length                                46
    0000                    status                              completed
    0000                    segment                               21
    1111                       type                               description  #new row
    1111                     tp                                   0.1
    1111                     x                                    56
    2222                     type                                description  #new row
    2222                     point                                23.01
    2222                     x                                    50
    2222                     y                                    40

i know about loc[] or iloc[] but i can not use them as i dont know the exact index or position.I am new to pandas. Is there a way to do it? any help will be ppreciated

Answer 1

new = df.groupby('id', as_index=False).first().assign(name='type', value='description  #new row')
df = pd.concat([new, df]).sort_values('id')

print(df)

Prints:

     id     name                  value
0  0000     type  description  #new row
0  0000   length                     46
1  0000   status              completed
2  0000  segment                     21
1  1111     type  description  #new row
3  1111       tp                    0.1
4  1111        x                     56
2  2222     type  description  #new row
5  2222    point                  23.01
6  2222        x                     50
7  2222        y                     40

Note: this assumes that column id is sorted before.