EpistasisLab / EpistasisLab/tpot

TPOT MDR - no usage of MDR?

Open
#1,083 7 comments 0 reactions 0 assignees View on GitHub
question
Dominant language
Jupyter Notebook
Stars
10.1k
Forks
1.6k
PR merge metrics
No merged PRs in 30d

Description

I've been running TPOT with the TPOT MDR configuration, and in multiple runs I see no usage of any of the MDR functionalities - MDR, ReliefF, Surf, etc. For example, the following pipeline came from a run with a population of 250 for 25 generations:

exported_pipeline = make_pipeline(
make_union(
FunctionTransformer(copy),
StackingEstimator(estimator=ElasticNetCV(l1_ratio=1.0, tol=0.01))
),
ElasticNetCV(l1_ratio=0.15000000000000002, tol=0.01)
)

I have done similar runs with 100 individuals for 15 generations using a different random seed, and I see very similar results, but I've never seen one of the MDR functions appear in a pipeline.

Do I need to supply an additional argument that I'm not aware of to enable MDR and skrebate functions?

Versions:

tpot = 0.11.5
skrebate = 0.6
scikit-learn = 0.22.1
scikit-mdr = 0.4.4
scipy = 1.4.1
numpy = 1.17.5

Contributor guide

No contributing guide indexed for this repository

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.