module 'pandas' has no attribute 'read_csv

Question

import pandas as pd

df = pd.read_csv('FBI-CRIME11.csv')

print(df.head())

Running this simple code gives me the error:

Traceback (most recent call last):
  File "C:/Users/Dita/Desktop/python/lessons/python.data/csv.py", line 1, in <module>
    import pandas as pd
  File "C:\python\lib\site-packages\pandas-0.19.1-py3.5-win-amd64.egg\pandas\__init__.py", line 37, in <module>
    import pandas.core.config_init
  File "C:\python\lib\site-packages\pandas-0.19.1-py3.5-win-amd64.egg\pandas\core\config_init.py", line 18, in <module>
    from pandas.formats.format import detect_console_encoding
  File "C:\python\lib\site-packages\pandas-0.19.1-py3.5-win-amd64.egg\pandas\formats\format.py", line 33, in <module>
    from pandas.io.common import _get_handle, UnicodeWriter, _expand_user
  File "C:\python\lib\site-packages\pandas-0.19.1-py3.5-win-amd64.egg\pandas\io\common.py", line 5, in <module>
    import csv
  File "C:\Users\Dita\Desktop\python\lessons\python.data\csv.py", line 4, in <module>
    df = pd.read_csv('FBI-CRIME11.csv')
AttributeError: module 'pandas' has no attribute 'read_csv'

Can you rerun `pip install pandas` and try again? – Nov 11 '16 at 19:17 — , Nov 11 '16 at 19:17

score 76 · Accepted Answer · answered Nov 11 '16 at 19:16

76

Try renaming your csv.py to something else, like csv_test.py. Looks like pandas is being confused about what to import.

answered Nov 11 '16 at 19:16

AKX

152,115
15
115
172

You may also need to remove any lingering `csv.pyc` file. – chadlagore Jun 04 '18 at 18:55
1

Any elaboration as to why that happened? – meow Feb 19 '19 at 16:22
5

@meow There is a builtin python module called `csv`, but having an user module by the same name has precedence and Pandas tries to import it. – AKX Feb 20 '19 at 05:14
Where should I look for this csv.py file? in the folder of the environment am using? (FYI- am using a virtual environment) – Vikrant Arora Aug 16 '20 at 16:06

tommmm · Answer 2 · 2023-02-24T03:07:07.783

15

Make sure you don't have a file called pandas.py in the directory you are executing your python file in.

More info in Euguene's answer in this post.

edited Feb 24 '23 at 03:07

answered Apr 02 '21 at 01:56

tommmm

182
1
4

1

I actually had my script named pandas.py because I was testing pandas functionality, good tip. – Anonymous Nov 01 '21 at 15:29
I also named my file pandas.py haha seems like an occuring mistake! thanks for the tip – Fawzy Hegab Jul 14 '22 at 20:54

Atul Balaji · Answer 3 · 2020-03-21T04:07:59.180

5

I had checked for the presence of csv.py and made sure there was no file of this name. I also tried pip uninstall pandas and then pip install pandas. I still got the same error.

What worked for me was: pip-autoremove.

First install it using pip install pip-autoremove.
Then, remove pandas using pip-autoremove pandas -y.
Next, reinstall it using pip install pandas.

The reason why this is necessary is that sometimes, when using uninstall, the package folder may still be present.

edited Mar 21 '20 at 04:07

answered Mar 01 '20 at 14:04

Atul Balaji

804
4
15
29

I tried this approach. but I end up getting this : ```pkg_resources.DistributionNotFound: The 'panadas' distribution was not found and is required by the application``` – julian joseph Oct 16 '20 at 11:37

score 3 · Answer 4 · answered Sep 26 '20 at 13:18

I'm using Jupiter notebook and in my case I have imported and used pandas as below.

import pandas as pd
df = pd.read_csv('canada_per_capita_income.csv')

but it keeps throwing this error

DataFrame' object has no attribute 'read_csv

then i restart the kernel and rename the csv file to

income.csv

it solve my issue. Hope this might help someone

score 3 · Answer 5 · answered Jul 25 '21 at 07:14

3

You need to make sure there are no files named pandas.py, numpy.py or matplotlib.py

answered Jul 25 '21 at 07:14

Harshal Deore

1,050
1
11
11

module 'pandas' has no attribute 'read_csv

5 Answers5

Linked