It seems that when running
pip install . --user on a project, it installs all dependencies completely outside the virtualenv. I noticed this while following the installation instructions of keras-retinanet:
1. Clone this repository. 2. In the repository, execute pip install . --user. Note that due to inconsistencies with how tensorflow should be installed, this package does not define a dependency on tensorflow as it will try to install that (which at least on Arch Linux results in an incorrect installation). Please make sure tensorflow is installed as per your systems requirements.
After doing these steps, I added the virtualenv path to my Pycharm IDE project preferences, but it didn’t recognize any of the dependencies. When running
pip list inside my virtualenv, none of the dependencies showed up, though I do see them in
I tried to “undo” this via
pip uninstall . --user , but that gave an error saying that the uninstall subcommand didn’t recognize the --user flag. How can I clean this up?
Should trying to
pip install --user while in a virtualenv just throw an error? It seems misleading in that it appears to work fine, but as far as I can tell the virtualenv is completely ignored, and everything escapes into the global state.