awslabs/gluonts

Add exogenous features to `statsforecast` model wrappers

Open

#2,410 opened on Nov 2, 2022

View on GitHub
 (0 comments) (0 reactions) (0 assignees)Python (3,888 stars) (753 forks)batch import
enhancementgood first issue

Description

Description

The wrapper for statsforecast models added in #2360 does not use any exogenous features (only the past target is passed on to the model). It would be good to allow other standard fields from data entries, like at least feat_dynamic_real (or the past_ one) and feat_static_real, to be used.

Reference

See statsforecast's documentation on exogenous features: https://nixtla.github.io/statsforecast/examples/exogenous.html

Contributor guide

Tech stack
python
Domain
datamachine learning
Issue type
feature
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.
1-2 days
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
Pythontime series fundamentals
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
50
Research direction
Examine the existing statsforecast wrapper implementation added in PR #2360 to understand how the model wrapper currently works. Identify the specific wrapper class and its method that passes data to the model. Then, modify that method to accept additional arguments for exogenous features like feat dynamic real and feat static real, following the statsforecast documentation at https://nixtla.github.io/statsforecast/examples/exogenous.html. Ensure compatibility with the existing data structure.