Expand description
§use-ml-feature
Feature metadata primitives for RustUse machine-learning workflows.
§Experimental
use-ml-feature is experimental while use-ml remains below 0.3.0.
§Example
use use_ml_feature::{MlFeatureKind, MlFeatureName, MlFeatureRole};
let name = MlFeatureName::new("sepal_width")?;
let kind: MlFeatureKind = "numeric".parse()?;
assert_eq!(name.as_str(), "sepal_width");
assert_eq!(kind, MlFeatureKind::Numeric);
assert_eq!(MlFeatureRole::Input.as_str(), "input");§Scope
- Feature names, identifiers, kinds, roles, and sources.
- Transform, encoding, scaling, missing-value, and drift-status labels.
- Lightweight validation for ASCII-safe feature names.
§Non-goals
- Performing feature engineering, transformations, or pipeline orchestration.
- Prompt variables, message fields, context-window fields, or tool-call arguments.
§License
Licensed under either Apache-2.0 or MIT.