Dataset energy shifting

[chrisiacovella]

Pull Request Summary

This PR adds in the ability to shift the energies of systems in a dataset (after self-energies have been removed). This gives the option to shift by the minimum value (making all values positive), the maximum value (making all values negative), or the mean (centering the distribution at zero).

Key changes

Notable points that this PR has either accomplished or will accomplish.

• the toml file now allows a user to specify a shifting scheme, triggering this operation during the per-datapoint preprocessing operations.

Associated Issue(s)

• Dataset shifting #387

Pull Request Checklist

• Issue(s) raised/addressed and linked
• Includes appropriate unit test(s)
• Appropriate docstring(s) added/updated
• Appropriate .rst doc file(s) added/updated
• PR is ready for review

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 91.01%. Comparing base (dfa842b) to head (8d20cd7).

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[Copilot]

Pull Request Overview

This PR adds support for energy shifting to the training pipeline, allowing energies to be shifted by their minimum, maximum, or mean values to potentially improve training stability and speed.

Key changes:

• Introduces a shift_energies parameter with support for 'min', 'max', and 'mean' shifting modes
• Implements energy shifting logic in the dataset processing pipeline after self-energy removal
• Adds comprehensive test coverage for all three shifting modes

Reviewed Changes

Copilot reviewed 7 out of 7 changed files in this pull request and generated 7 comments.

Show a summary per file

File	Description
modelforge/train/parameters.py	Adds ShiftEnergiesMode enum and shift_energies field to training parameters
modelforge/train/training.py	Passes shift_energies parameter to datamodule setup
modelforge/dataset/dataset.py	Implements energy shifting logic in _per_datapoint_operations and adds parameter to method signatures
modelforge/tests/test_dataset.py	Adds comprehensive test for all three energy shifting modes
modelforge/tests/conftest.py	Updates datamodule factory to accept shift_energies parameter
docs/training.rst	Documents the new shift_energies feature
docs/datasets.rst	Documents energy shifting in the dataset operations section

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[chrisiacovella]

CI is failing due to failing to install the reference version of sake. I've commented this out, but will need revise the testing to use static information (as we've done for schnet, etc.). #389

[Copilot]

Pull Request Overview

Copilot reviewed 12 out of 12 changed files in this pull request and generated 5 comments.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.