I am trying to turn multiple columns into a single column, without the "null" ones and keep the identifier of each row
| Identifier | Column1 | Column2 | Column3 | Column4 |
|---|---|---|---|---|
| 1 | Dog | Cow | Sheep | Dinosaur |
| 2 | Dog | Pig | ||
| 3 | Bull | Cow | Elephant |
I want the new 2 columns like this, the original dataframe might have lots of columns, 20, 30, maybe more.
| Identifier | Var |
|---|---|
| 1 | Dog |
| 1 | Cow |
| 1 | Sheep |
| 1 | Dinosaur |
| 2 | Dog |
| 2 | Pig |
| 3 | Bull |
| 3 | Cow |
| 3 | Elephant |
>Solution :
You can use melt() for this
df.melt(id_vars = 'Identifier')[['Identifier', 'value']].dropna().sort_values('Identifier')