Pandas import error

Question

I tried installing pandas using easy_install and it claimed that it successfully installed the pandas package in my Python Directory.

I switch to IDLE and try import pandas and it throws me the following error -

Traceback (most recent call last): File "<pyshell#0>", line 1, in <module> import pandas File "C:\Python27\lib\site-packages\pandas-0.12.0-py2.7-win32.egg\pandas\__init__.py", line 6, in <module> from . import hashtable, tslib, lib File "numpy.pxd", line 157, in init pandas.hashtable (pandas\hashtable.c:20282) ValueError: numpy.dtype has the wrong size, try recompiling

Please help me diagnose the error.

FYI: I have already installed the numpy package

Answer 1

Maybe you interrupted pandas install , retry using pip :

First install pip (if you haven't done it already) :

easy_install pip

then reinstall pandas:

pip install pandas --upgrade

Hope it helps

Answer 2

You know that output error you got when you tried running @nipun-batra's script?

Well, you got it because you have to first:

import platform

before you can run:

platform.platform()  

I know this because I--about 10 minutes ago--got the same error when trying to run the same script. The difference is that I--an absolute beginner--figured out our problem after a quick trip over to google. (Man, they let you search for anything over there!)

This, when coupled with your follow-up appeal exactly two months after your initial posting, suggests to me that you would prefer to minimize--as much as possible--the usual hardship associated with owning and operating your own computer-machine-thingy.

As a result, with respect to your initial IDLE/pandas issue, your best best bet is to forget about messing around with easy_install, etc. Instead, go head on down to Continuum Analytics and pick up your very own (free) copy of Anaconda, which has got more packages than you can shake a stick at! (Including, I might add, pandas, numpy, scipy, statsmodels, matplotlib, IPython, and many more). And the best part is that it all comes bundled together as a single easy-to-download file. Trust me, it will save you a lot of headaches if you just download everything all at once.

Hope this helps!