Closed oditorium closed 8 years ago
Hi, Stefan, I am sorry about the trouble, and I don't know what the issues could be (yet).
Personally, I am not testing on Ubuntu locally (only have a few Macs, and a few RedHat and CentOS machines, but I am only testing locally on a Mac). However, as far as I know, the continuous integration tests (via Travis) are running on Ubuntu 14.04 LTS Server Edition 64 bit as described here. So, it should technically work on Ubuntu as well ... or that's what I thought. I am currently attending a conference in Toronto and can't test it on any of my other linux machines locally, but on my MacBook it installs fine using the .egg
via pip (see screenshot).
I'll try it locally on my other linux machines next week when I am back; meanwhile, if you have any ideas about what could cause this error, please let me know, I am more than happy to fix this (if I can) :).
PS: Have you tried to fork the repo (or download the zip) and install it via python setup.py install
from its root directory? One thing that I could spontaneously think of would be the 5k MNIST dataset, which is quite large, maybe there's a problem in the new Ubuntu version with parsing this relatively large .py file, which may cause the "Killed" result when it takes too long?
Hi, Stefan, I just created conda packages for Mac, Linux, and Windows. Maybe this could help with the problem? You can install it via
conda install -c rasbt mlxtend
Please let me know if this is helpful and solves this issue :)
I recently installed it on an ubuntu machine using the conda dist and it worked fine for me. I am closing this issue now (in hope it is resolved?), but in case you still have troubles, please let me know, and I'd be happy to look into it further!
Hi.
Below commands that I run on a clean install of Ubuntu 16.04 / x64 (on DigitialOcean)
Preparation:
Pip3 install (note 'Killed')
Install from github