Filtro por fila contiene pandas
In [3]: df[df['ids'].str.contains("ball")] # removes all rows where 'ball' not in row['ids']
Out[3]:
ids vals
0 aball 1
1 bball 2
3 fball 4
Nasty Nightingale