alteryx/evalml

Linear model throws convergence error on city_rev.csv

Open

#2,763 opened on Sep 9, 2021

View on GitHubContributor guide
 (2 comments) (0 reactions) (0 assignees)Python (93 forks)auto 404
buggood first issueperformance

Repository metrics

Stars
 (852 stars)
PR merge metrics
 (PR metrics pending)

Description

Preview of the error: image

We've noticed this for several of our perf testing datasets. One in particular was regress.csv Cloudwatch link here The warning being printed:

2021-09-09T17:33:44.934-04:00 | //.pyenv/versions/3.8.3/lib/python3.8/site-packages/sklearn/linear_model/_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):
-- | --
  | 2021-09-09T17:33:44.934-04:00 | STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.
  | 2021-09-09T17:33:44.934-04:00 | Increase the number of iterations (max_iter) or scale the data as shown in:
  | 2021-09-09T17:33:44.934-04:00 | https://scikit-learn.org/stable/modules/preprocessing.html
  | 2021-09-09T17:33:44.934-04:00 | Please also refer to the documentation for alternative solver options:
  | 2021-09-09T17:33:44.934-04:00 | https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression
  | 2021-09-09T17:33:44.934-04:00 | n_iter_i = _check_optimize_result(

This could be in relation to the Elastic Net models we have.

Contributor guide

Research direction
Investigate the convergence warning for logistic regression in sklearn. Consider increasing max iter or scaling the data using StandardScaler. Check if the issue is related to Elastic Net models and reproduce with city rev.csv.
Tech stack
pythonscikit learn
Domain
backenddatamachine learning
Issue type
Bug
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
2
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.
Active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Clear
Prerequisites
Pythonscikit learn
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
70