0

in order to use a specific implementation, I need to install a variety of older versions of packages in Python3.4. I made a virtualenv with 3.4.

Everything goes smoothly until I try to install pandas 0.21.0 with pip, because it throws errors regarding numpy (which should be 1.13.3 for the implementation). I'm not going to post the error because it is very long, but pandas tries to install the latest numpy, which doesn't seem to support 3.4.

How could I get around this and make pandas install a specific numpy version or overgo the installation of numpy? I'm using Linux btw.

Thanks!

hantyumi_
  • 1
  • 1
    Why don't you install required numpy version manually, before installing the required version of pandas? – h4z3 Jan 13 '20 at 11:37
  • Even if I do that, pandas still tries to install the newest numpy version :( – hantyumi_ Jan 13 '20 at 13:13
  • Does this answer your question? [Upgrade package without upgrading dependencies using pip?](https://stackoverflow.com/questions/2861183/upgrade-package-without-upgrading-dependencies-using-pip) – h4z3 Jan 13 '20 at 16:20

0 Answers0