Reputation: 96398
Pandas
has a number of dependencies, e.g matplotlib
, statsmodels
, numexpr
etc.
Say I have Pandas
installed, and I update many of its dependencies, if I don't update Pandas, could I run into any problems?
Upvotes: 0
Views: 326
Reputation: 102902
If your version of pandas
is old (i.e., not 0.13.1), you should definitely update it to take advantage of any new features/optimizations of the dependencies, and any new features/bug fixes of pandas
itself. It is a very actively-maintained project, and there are issues with older versions being fixed all the time.
Of course, if you have legacy code that depends on an older version, you should test it in a virtualenv with the newer versions of pandas
and the dependencies before updating your production libraries, but at least in my experience the newer versions are pretty backwards-compatible, as long as you're not relying on buggy behavior.
Upvotes: 3