pandas - A value is trying to be set on a copy of a slice from a DataFrame

Asked Nov 24 '17 at 07:57

Active Nov 24 '17 at 07:57

Viewed 44 times

I know this has been asked before but my case is slightly different . I am using the below .

    train_mask = df_train['category_count'] == 3

    df_train.loc[train_mask]['cat1'] = df_train.loc[train_mask]['categories'][0[0]
    df_train.loc[train_mask]['cat2'] = df_train.loc[train_mask]['categories'][0][1]
    df_train.loc[train_mask]['cat3'] = df_train.loc[train_mask]['categories'][0][2]


where for eg : df_train['categories'] is a list  : ["Shirt","Trousers","Shoes"]

I am using the .loc method inspite of which i keep getting the below error .Can someone please explain why .

C:\Users\sraghunath\AppData\Local\Continuum\Anaconda3\lib\site-packages\ipykernel_launcher.py:1: SettingWithCopyWarning: A value is trying to be set on a copy of a slice from a DataFrame. Try using .loc[row_indexer,col_indexer] = value instead

asked Nov 24 '17 at 07:57

user1411837

1,494
6
27
41

I think you can check [this](https://stackoverflow.com/q/20625582/2901002) – jezrael Nov 24 '17 at 07:59
2

So maybe need `df_train.loc[train_mask, 'cat1'] = df_train.loc[train_mask, 'categories'][0][0]`, but not sure with last `[0][0]`. – jezrael Nov 24 '17 at 08:00
If it does not work or some problem, please add some sample data to question. – jezrael Nov 24 '17 at 08:01
@jezrael - Perfect...guess i can't accept it as an answer since its a comment..but yes...many thanks for fixing my issue. – user1411837 Nov 24 '17 at 08:03
It is dupe, so I write only comment :( – jezrael Nov 24 '17 at 08:04

pandas - A value is trying to be set on a copy of a slice from a DataFrame

0 Answers0