Python Pandas: AttributeError: 'str' object has no attribute 'loc'

Question

Working with dataframe df:

Count
1
2
3
4
5

Want to add second column, that categorizes everything above 3 as '4+' - needed output:

Count | Category
1        1
2        2
3        3
4        4+
5        4+

This is my code:

df['Category'] = df['Count']
df = df.loc[df['Count'] > 3, 'Category'] = '4+'

And I get this error:

AttributeError: 'str' object has no attribute 'loc'

FYI - you can also use `df['Category'] = np.where(df['Count'] < 4, df['Count'], '4+')` — Zero, Sep 20 '17 at 17:42

score 5 · Accepted Answer · answered Sep 20 '17 at 17:31

5

Just go with

df['Category'] = df['Count']
df.loc[df['Count'] > 3, 'Category'] = '4+'

answered Sep 20 '17 at 17:31

Vaishali

thank you - do you mind explainin, what I did wrong? – jeangelj Sep 20 '17 at 17:31
I just tried without the "df=" and I get the same error – jeangelj Sep 20 '17 at 17:32
Why wouldn't I use the df=? – jeangelj Sep 20 '17 at 17:35
2

You are already assigning using loc when you do df.loc[df['Count'] > 3, 'Category'] = '4+'. When you add df = ...., you are trying to assign twice in the same statement, which obviously wouldn't work – Vaishali Sep 20 '17 at 17:36
Is this answer actually working for you @jeangelj? I tried it and got Error `ValueError: invalid literal for int() with base 10: '4+'` – Cedric Zoppolo Sep 20 '17 at 18:26
If the column `"Count"` has integers you should do `df['Category'] = df['Count'].apply(str)` first as I stated in my answer. – Cedric Zoppolo Sep 20 '17 at 18:29

score 1 · Answer 2 · answered Sep 20 '17 at 18:13

1

You can try out with:

import pandas as pd
df = pd.DataFrame({"Count": [1,2,3,4,5]})
df["Category"] = df["Count"].apply(str)
df["Category"][df['Count'] > 3] = "4+"

Output would be:

>>> df
   Count Category
0      1        1
1      2        2
2      3        3
3      4       4+
4      5       4+

answered Sep 20 '17 at 18:13

Cedric Zoppolo

It throws SettingWithCopyWarning: A value is trying to be set on a copy of a slice from a DataFrame – Vaishali Sep 20 '17 at 18:32
Did you try the only the example I stated or you used your own dataframe? If you dataframe has numbers this should work. – Cedric Zoppolo Sep 20 '17 at 18:52
Please go through this link to understand settingwithcopy warning and why should loc be used. https://stackoverflow.com/questions/20625582/how-to-deal-with-settingwithcopywarning-in-pandas – Vaishali Sep 20 '17 at 18:59
Well, that's strange. I tried under Python 2.7 and 3.3 on [pythonanywhere.com](http://www.pythonanywhere.com) and got no warnings from my option. – Cedric Zoppolo Sep 20 '17 at 19:21

2 Answers2