Submodules aren't implicitly imported, and must be explicitly declared, but I'm making an explicit call to the pd.Series
submodule, aren't I?
Regardless, shouldn't import pandas as pd
allow for pd.Series
to be called? The following code works flawlessly in iPython, but fails when executed from a script.
#!/usr/bin/env/python2.7
# -*- coding: utf-8 -*-
import pandas as pd
import numpy as np
counts = pd.Series([632, 1638, 569, 115], index=["Firmicutes", "Proteobacteria", "Actinobacteria", "Bacteroidetes"])
Results in:
tyler@machine ~/src/stats $ python pandas.py
Traceback (most recent call last):
File "pandas.py", line 3, in <module>
import pandas as pd
File "/home/tyler/src/stats/pandas.py", line 6, in <module>
counts = pd.Series([632, 1638, 569, 115], index=["Firmicutes", "Proteobacteria", "Actinobacteria", "Bacteroidetes"])
AttributeError: 'module' object has no attribute 'Series'
Where have I gone wrong?
The issue is that you've called your module
pandas
. Call it something else. And don't forget to delete thepandas.pyc
generated onimport pandas
or else it will keep failing.I had to same issue with the error message "'module' object has no attribute '_tseries'", I have missed to install python-dateutil package
The solution:
with python2.6, to install pandas 0.6.1, you have to install numpy 1.6.1 and python-dateutil like 2.6.1
Good luck ;)
series in ipython not python