+- Python Forum (https://python-forum.io)
+-- Forum: Python Coding (https://python-forum.io/forum-7.html)
+--- Forum: Data Science (https://python-forum.io/forum-44.html)
+--- Thread: Lime for regression interpretability (/thread-24414.html)
Lime for regression interpretability - Scott - Feb-12-2020
Hi everyone,
I am trying to pip install lime for regression interpretability but am getting the following error in Anaconda prompt:
ERROR: Could not find a version that satisfies the requirement lime (from versions: none)
ERROR: No matching distribution found for lime
The package is a PyPl file tho:
https://pypi.org/project/lime/
Has anyone come across this issue before?
Thanks
Scott
RE: Lime for regression interpretability - Larz60+ - Feb-12-2020
the author hasn't even supplied
author's GitHub repository has more documentation here: https://github.com/marcotcr/lime
Have you tried installing with conda?
RE: Lime for regression interpretability - Scott - Feb-13-2020
Yes I have tried conda install lime, I get a different error but it does not work. I also tried the setup file in the link you provided, by running the code in jupyter notebook but got an error.
RE: Lime for regression interpretability - Larz60+ - Feb-13-2020
Can't find a wheel, but there is a tar file here: https://files.pythonhosted.org/packages/b5/e0/60070b461a589b2fee0dbc45df9987f150fca83667c2f8a064cef7dbac6b/lime-0.1.1.37.tar.gz
I installed it on my Linux system, running python 3.8.1 without a hitch using pip.
which version of python is being used by your anaconda system, on which OS?
RE: Lime for regression interpretability - jefsummers - Feb-14-2020
Backing up, are you sure you need Lime and can't get what you want from Scikit-learn?
RE: Lime for regression interpretability - Scott - Feb-14-2020
I cannot find anything in Scikit-learn that shows variables predicative power. Anyone else know?
On another note I can install lime on my personal computer. I think my work python has been compromised by me trying to get round the firewall and installing more packages.