I’m a new prodigy user and I think the tool looks super cool.
Unfortunately, I’m having issues even loading it.
I get a:
ImportError: Something is wrong with the numpy installation. While importing we detected an older version of numpy in [‘path/to/anaconda/package’]. One method of fixing this is to repeatedly uninstall numpy until none is found, then reinstall this version.
Whenever I try to run a recipe. Have tried to uinstall and reinstall and do it on and off environements and neither seems to work.
Thanks in advance for your time,
Hi and welcome! From looking at the error, I think the “good” news is that this is most likely not a problem with Prodigy and probably related to your conda environment(s). For some reason, you’ve ended up with conflicting version of numpy, which is a dependency of Prodigy, but probably also most other packages you use.
How did you install Prodigy? Did you use a fresh virtual environment or did you install it into your system environment? And what happens if you follow numpy’s advice in the error message and uninstall it until all conflicting installations are gone, and then reinstall it? If you search for the message on Stack Overflow, you’ll also find various threads that basically recommend the same thing.
I guess I skipped too fast over the repeatedly uninstall part
Thanks a bunch!