While trying to use groupby to concatenate strings, i stumbled upon this code:
df_g = df.groupby([\'user-id\'])[\'track\'].apply(lambda x: "{%s}" %