This is fork with updates enabling build of 0.23.X in python 3.11 The objcative of this fork is to maintain the performance of the mixed gaussian estimator used to extract liver reference values in aPROMISE. Later versions of scikit learn (>1.0.0) have updated parameter initializations using a normal distribution instead of a uniform, changing the convergent estimate in multiple cases.
Deploy package to aws.
Build wheel using
$ python setup.py bdist_wheel -d dist/
from the base folder.
Upload new package version to release to aws. If version is x.y.z, upload wheel to aws. For location see pattern in image-analysis.