From lists in a column to rows

Question

I have this dataframe

Node                  TLW                            
1                  [2, 22, 3]                           
2                     [12]                              
3                    [2,43]                             
4                     [3]                             
5                     [11]  

                        

I would like to have something like this

Node
1
2
3
4
5
22
12
43
11

Could you please tell me how to get it? I would try to use for loop in the list and then append to my dataframe, checking for duplicates. This would be my approach, but I am still having difficulties in using for loop here. I was thinking of using explode but the output would be not what I am looking for, as the (distinct) numbers (or strings) should be in the column Node, not in TLW.

Answer 1

Method 1

One way is to use apply+lambda to merge the Node and the TLW column into a single list. Then use explode and take a unique(). Post that recreate a dataframe with single column Node

d = {'Node': {0: 1, 1: 2, 2: 3, 3: 4, 4: 5},
     'TLW': {0: [2, 22, 3], 1: [12], 2: [2,43], 3: [3], 4: [11]}}
df = pd.DataFrame(d)

nodes = df.apply(lambda x: [x['Node']]+ x['TLW'], axis=1).explode().unique()
new_df = pd.DataFrame(nodes, columns=['Node'])
print(new_df)

  Node
0    1
1    2
2   22
3    3
4   12
5   43
6    4
7    5
8   11

---

Method 2

Another way would be using numpy's np.unique after df.explode -

import numpy as np

d = {'Node': {0: 1, 1: 2, 2: 3, 3: 4, 4: 5},
     'TLW': {0: [2, 22, 3], 1: [12], 2: [2,43], 3: [3], 4: [11]}}
df = pd.DataFrame(d)


new_df = pd.DataFrame(np.unique(df.explode('TLW').values), columns=['Nodes'])
print(new_df)

  Nodes
0     1
1     2
2     3
3     4
4     5
5    11
6    12
7    22
8    43