automl/auto-sklearn

Stress test repeated construction and fitting from same process

Open

#1,303 opened on Nov 17, 2021

View on GitHub
 (3 comments) (0 reactions) (0 assignees)Python (7,270 stars) (1,265 forks)batch import
Good first issuemaintenance

Description

From issue #1302, it appears autosklearn is a bit unstable when run many times in the same script, i.e. in a for loop.

for i in range(400):
    automodel = AutoSklearn(full_resources)
    automodel.fit(x, y)

We currently have no test for this and it would be good to see if we can reproduce the same connection refused error.

Contributor guide

Tech stack
pythonscikit learn
Domain
testingmachine learning
Issue type
test
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
3
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
1-3 hours
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
stale
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
clear
Prerequisites
Pythonpytestauto sklearn basics
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
60
Research direction
Investigate the connection refused error reported in issue #1302. Review existing test files in the repository, particularly those in the 'test' directory, for patterns on how to write stress tests. Write a test that instantiates and fits AutoSklearn models in a loop (e.g., 400 iterations) to see if the error reproduces. Pay attention to resource cleanup between iterations.