chore: release v1.3.3

[ArthurBernard]

[1.3.3] - 2026-05-11

Added

• sortino(X, rf, period, ...) evaluation metric in fynance.features.metrics, symmetric to sharpe
• directional_accuracy(y_true, y_pred) evaluation metric — percentage of correctly predicted return signs
• fynance.models.loss submodule with differentiable PyTorch loss functions: SharpeLoss, SortinoLoss, DirectionalAccuracyLoss, and base class BaseLoss
• Tests for new metrics and loss functions; overall test and docstring coverage above 80%

Changed

• _compute_returns factorised as a shared helper; _annual_volatility and _annual_downside_volatility use it — removes duplicated logic and double allocation in sortino()
• accuracy() simplified to a single numpy pass
• BaseLoss.__init__ precomputes rf / period to avoid per-forward division
• SharpeLoss uses std(correction=0) for consistency with the numpy sharpe metric

Fixed

• Reformat long import in test_allocation.py (ruff I001)

Removed

• Dead commented-out __add__ / __iadd__ stubs from LossSeries
• TODO.md untracked from git (already declared in .gitignore)

Test plan

• CI verte sur la branche
• pyproject.toml et CHANGELOG en phase sur v1.3.3

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

✅ All modified and coverable lines are covered by tests.
❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

🚀 New features to boost your workflow:

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