Update to Code about omnianomaly HOT 3 CLOSED

Yes, we do encourage people to create to run the code in an isolated environment. That's why virtualenv is recommended in README. You can use it by

virtualenv <env_name>
source <env_name>/bin/activate
pip install -r requirements.txt

And there are equivalent commands for conda. You can use either way to build an isolated python environment.

from omnianomaly.

1. Our code is compatible with TensorFlow 1.12.0 (we have fixed it in requirements.txt). Sorry, we cannot update it to a more recent version, due to the dependency on TensorFlow-probability 0.5.0.
2. The error introduced by test_labels directory name has been fixed in a recent commit.

from omnianomaly.

Maybe just create a conda environment for everything. That way when people test the code they can do it in an isolated environment

from omnianomaly.