Hi Arcpy Users,
Just want to inquire how can I properly install Pandas python package in ArcPy. I am having a hard time installing it. I just wanted to get the pandas to work so I can work easier with my CSV.
Is this has something to do with my pip installation?
I am using this command:
>>> pip install pandas
and it returns this message:
Microsoft Windows [Version 6.1.7601]
Copyright (c) 2009 Microsoft Corporation. All rights reserved.
C:\Users\lpalao>pip install pandas
Traceback (most recent call last):
File "C:\Python27\ArcGIS10.2\Scripts\pip-script.py", line 9, in <module>
load_entry_point('pip==6.0.8', 'console_scripts', 'pip')()
File "build\bdist.win32\egg\pkg_resources.py", line 339, in load_entry_point
File "build\bdist.win32\egg\pkg_resources.py", line 2470, in load_entry_point
File "build\bdist.win32\egg\pkg_resources.py", line 2184, in load
File "C:\Python27\ArcGIS10.2\lib\site-packages\pip-6.0.8-py2.7.egg\pip\__init__.py", line 13, in <
from pip.utils import get_installed_distributions, get_prog
File "C:\Python27\ArcGIS10.2\lib\site-packages\pip-6.0.8-py2.7.egg\pip\utils\__init__.py", line 22
, in <module>
from pip._vendor import pkg_resources, six
File "C:\Python27\ArcGIS10.2\lib\site-packages\pip-6.0.8-py2.7.egg\pip\_vendor\__init__.py", line
61, in load_module
t__.py", line 79, in <module>
s.py", line 274, in <module>
s.py", line 373, in Specifier
re.VERBOSE | re.IGNORECASE,
File "C:\Python27\ArcGIS10.2\lib\re.py", line 190, in compile
return _compile(pattern, flags)
File "C:\Python27\ArcGIS10.2\lib\re.py", line 242, in _compile
raise error, v # invalid expression
sre_constants.error: nothing to repeat
Then when I import pandas (>>> import pandas) it returns the message below:
>>> import numpy
>>> import pandas
numpy.ufunc has the wrong size, try recompiling
Traceback (most recent call last):
File "<interactive input>", line 1, in <module>
File "C:\Python27\ArcGIS10.2\lib\site-packages\pandas-0.15.2-py2.7-win32.egg\pandas\__init__.py", line 7, in <module>
from . import hashtable, tslib, lib
File "pandas\src\numpy.pxd", line 865, in init pandas.hashtable (pandas\hashtable.c:22988)
ValueError: numpy.ufunc has the wrong size, try recompiling
An alternate site, but here are the requirements from this page
I don't use it, but I have installed other modules from this site with no problem
I found that to successfully install pandas for use with ArcGIS 10.3 I had to install a newer version of numpy (upgrade from 1.7 to 1.9). In doing that, I had to make sure to use pip in a command window with administrative permissions.
If you are ok with using an older version of pandas, the installation process I used was relatively simple:
I installed pandas by running an executable installation file (built with distutils) downloaded from pypi.python.org via download links on the pandas homepage. The key for me was choosing an older version of pandas that was built using the same version of NumPy that I was already running.
To find out what version of numpy you are running, import NumPy and then type:
Although pandas 15.2 is the latest release, it (the executable installation file) requires NumPy 1.8.1. I'm running ArcMap 10.2 which comes with NumPy 1.7.1. I browsed through the older releases of pandas (listed on the right hand pandas homepage) until I found a release that was based on NumPy 1.7.1, which happened to be pandas 12.0. It was only later that I realized that the pandas 13.0 and 13.1 were also based on NumPy 1.7.1, but I didn't bother to change it.
Although python-dateutil and pytz are listed as pandas 15.2 dependencies, I think they aren't for pandas 12.0 since I didn't bother installing them.
I hope that helps!