AttributeError: module object has no attribute “Se

问题:

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?

回答1:

The issue is that you've called your module pandas. Call it something else. And don't forget to delete the pandas.pyc generated on import pandas or else it will keep failing.

回答2:

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

pip install numpy==1.6.1
pip install python-dateutil==2.6.1
pip install --force-reinstall pandas==0.6.1

Good luck ;)

回答3:

series in ipython not python

try 

  $ ipython
  import pandas as pd
  import numpy as np

  counts = pd.Series([632, 1638, 569, 115], index=["Firmicutes", "Proteobacteria", "Actinobacteria", "Bacteroidetes"])