I was having the same problem now with Python 3.4.3.
I was using pandas-0.18.0.
Upgrading (using pip) solved the issue for me:
[sudo] pip install --upgrade pandas
The final result of the upgrade:
Successfully installed numpy-1.13.3 pandas-0.21.0 python-dateutil-2.6.1 pytz-2017.3 six-1.11.0
After this, the issue was gone!