pub trait RegressionModel<Input, Output>: OptimizableModel<Input, Output> {
// Required methods
fn mse(&self, x: &Input, y: &Output) -> Result<f64, ModelError>;
fn rmse(&self, x: &Input, y: &Output) -> Result<f64, ModelError>;
fn r2(&self, x: &Input, y: &Output) -> Result<f64, ModelError>;
fn compute_metrics(
&self,
x: &Input,
y: &Output,
) -> Result<RegressionMetrics, ModelError>;
}Expand description
A trait defining common metrics and evaluation methods for regression models.
This trait should be implemented by any model that performs regression tasks, allowing standardized evaluation through common regression metrics.
§Type Parameters
Input- The type of the input data (features).Output- The type of the output data (target values).
§Methods
mse- Calculates the Mean Squared Error between predictions and actual values.rmse- Calculates the Root Mean Squared Error between predictions and actual values.r2- Calculates the coefficient of determination (R²) score.compute_metrics- Calculates a comprehensive set of regression metrics.
§Errors
Methods return Result<_, ModelError> to handle cases where metric calculation
might fail, such as with empty inputs or numerical issues.
Required Methods§
fn mse(&self, x: &Input, y: &Output) -> Result<f64, ModelError>
fn rmse(&self, x: &Input, y: &Output) -> Result<f64, ModelError>
fn r2(&self, x: &Input, y: &Output) -> Result<f64, ModelError>
fn compute_metrics( &self, x: &Input, y: &Output, ) -> Result<RegressionMetrics, ModelError>
Dyn Compatibility§
This trait is not dyn compatible.
In older versions of Rust, dyn compatibility was called "object safety", so this trait is not object safe.