pandas!=0.17.0,>=0.16.0
should do what you need, I think On Wed, Mar 9, 2016 at 3:40 PM, Tennessee Leeuwenburg < tleeuwenb...@gmail.com> wrote: > I'm not aware of a way to avoid 'cursed versions' of packages. For > example, pandas 0.17.0 has an issue with duplicate rows. This bug is fixed > in 0.17.1, but also the previous version was fine. I apologies if there is > logic to handle this situation. > > I would like to express this as a "latest version but not 0.17.0". This is > because 'latest' might mean different things depending where you are > getting your versions from (e.g. anaconda, pypi, OS packages). > > I don't know what to suggest, but I thought this was an interesting use > case and worth mentioning. > > > _______________________________________________ > Distutils-SIG maillist - Distutils-SIG@python.org > https://mail.python.org/mailman/listinfo/distutils-sig > >
_______________________________________________ Distutils-SIG maillist - Distutils-SIG@python.org https://mail.python.org/mailman/listinfo/distutils-sig
