Struct ModelComparison 

pub struct ModelComparison {
    pub model_a_id: String,
    pub model_b_id: String,
    pub performance_difference: Option<f64>,
    pub efficiency_difference: Option<f64>,
    pub statistical_significance: Option<bool>,
    pub recommendation: String,
}

Fields

model_a_id: String

model_b_id: String

performance_difference: Option<f64>

Relative change in the configured primary_metric from model_a to model_b, or None when at least one of them never recorded that metric (or the reference value is zero) – never a 0.0 standing in for “nothing was measured”.

efficiency_difference: Option<f64>

Mean of the relative training-time and model-size changes, or None when model_a has no recorded training time or model size.

statistical_significance: Option<bool>

Some(true)/Some(false) from a real Welch’s t-test over both models’ recorded primary_metric samples (see ModelComparator::test_statistical_significance), or None when there was not enough recorded history to run the test – never a fabricated constant.

recommendation: String

Trait Implementations

impl Debug for ModelComparison

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for ModelComparison

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for ModelComparison

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.