converge_analytics/packs/regression/
types.rs

1use converge_pack::gate::GateResult as Result;
2use serde::{Deserialize, Serialize};
3
4#[derive(Debug, Clone, Serialize, Deserialize)]
5pub struct RegressionInput {
6    pub records: Vec<Vec<f64>>,
7    pub weights: Vec<f64>,
8    pub bias: f64,
9}
10
11impl RegressionInput {
12    pub fn validate(&self) -> Result<()> {
13        if self.records.is_empty() {
14            return Err(converge_pack::GateError::invalid_input(
15                "At least one record required",
16            ));
17        }
18        let dim = self.weights.len();
19        if dim == 0 {
20            return Err(converge_pack::GateError::invalid_input(
21                "At least one weight (feature) required",
22            ));
23        }
24        for (i, record) in self.records.iter().enumerate() {
25            if record.len() != dim {
26                return Err(converge_pack::GateError::invalid_input(format!(
27                    "Record {} has {} features, expected {}",
28                    i,
29                    record.len(),
30                    dim
31                )));
32            }
33        }
34        Ok(())
35    }
36}
37
38#[derive(Debug, Clone, Serialize, Deserialize)]
39pub struct PredictedValue {
40    pub index: usize,
41    pub value: f64,
42}
43
44#[derive(Debug, Clone, Serialize, Deserialize)]
45pub struct RegressionOutput {
46    pub predictions: Vec<PredictedValue>,
47    pub mean_prediction: f64,
48    pub std_prediction: f64,
49    pub total: usize,
50}
51
52impl RegressionOutput {
53    pub fn summary(&self) -> String {
54        format!(
55            "Predicted {} values (mean: {:.3}, std: {:.3})",
56            self.total, self.mean_prediction, self.std_prediction,
57        )
58    }
59}
converge_analytics/packs/regression/types.rs

converge_analytics/packs/regression/
types.rs
