Struct RegressionOutput 

pub struct RegressionOutput {

    pub coefficients: Vec<f64>,
    pub std_errors: Vec<f64>,
    pub t_stats: Vec<f64>,
    pub p_values: Vec<f64>,
    pub conf_int_lower: Vec<f64>,
    pub conf_int_upper: Vec<f64>,
    pub r_squared: f64,
    pub adj_r_squared: f64,
    pub f_statistic: f64,
    pub f_p_value: f64,
    pub mse: f64,
    pub rmse: f64,
    pub mae: f64,
    pub std_error: f64,
    pub residuals: Vec<f64>,
    pub standardized_residuals: Vec<f64>,
    pub predictions: Vec<f64>,
    pub leverage: Vec<f64>,
    pub vif: Vec<VifResult>,
    pub n: usize,
    pub k: usize,
    pub df: usize,
    pub variable_names: Vec<String>,
    pub log_likelihood: f64,
    pub aic: f64,
    pub bic: f64,
}

Complete output from OLS regression.

Contains all coefficients, statistics, diagnostics, and residuals from an Ordinary Least Squares regression.

Example

let y = vec![1.0, 2.0, 3.0, 4.0, 5.0];
let x1 = vec![1.0, 2.0, 3.0, 4.0, 5.0];
let names = vec!["Intercept".to_string(), "X1".to_string()];

let result = ols_regression(&y, &[x1], &names).unwrap();
assert!(result.r_squared > 0.0);
assert_eq!(result.coefficients.len(), 2); // intercept + 1 predictor

// Save and load the model

Fields

coefficients: Vec<f64>

Regression coefficients (including intercept)

std_errors: Vec<f64>

Standard errors of coefficients

t_stats: Vec<f64>

t-statistics for coefficient significance tests

p_values: Vec<f64>

Two-tailed p-values for coefficients

conf_int_lower: Vec<f64>

Lower bounds of 95% confidence intervals

conf_int_upper: Vec<f64>

Upper bounds of 95% confidence intervals

r_squared: f64

R-squared (coefficient of determination)

adj_r_squared: f64

Adjusted R-squared (accounts for number of predictors)

f_statistic: f64

F-statistic for overall model significance

f_p_value: f64

P-value for F-statistic

mse: f64

Mean squared error of residuals

rmse: f64

Root mean squared error (prediction error in original units)

mae: f64

Mean absolute error of residuals

std_error: f64

Standard error of the regression (residual standard deviation)

residuals: Vec<f64>

Raw residuals (observed - predicted)

standardized_residuals: Vec<f64>

Standardized residuals (accounting for leverage)

predictions: Vec<f64>

Fitted/predicted values

leverage: Vec<f64>

Leverage values for each observation (diagonal of hat matrix)

vif: Vec<VifResult>

Variance Inflation Factors for detecting multicollinearity

n: usize

Number of observations

k: usize

Number of predictor variables (excluding intercept)

df: usize

Degrees of freedom for residuals (n - k - 1)

variable_names: Vec<String>

Names of variables (including intercept)

log_likelihood: f64

Log-likelihood of the model (useful for AIC/BIC calculation and model comparison)

aic: f64

Akaike Information Criterion (lower = better model, penalizes complexity)

bic: f64

Bayesian Information Criterion (lower = better model, penalizes complexity more heavily than AIC)

Trait Implementations

impl Clone for RegressionOutput

fn clone(&self) -> RegressionOutput

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for RegressionOutput

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for RegressionOutput

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl ModelLoad for RegressionOutput

fn load(path: &str) -> Result<Self>

Load a model from a file.

fn from_serialized(model: SerializedModel) -> Result<Self>

Load a model from an already-deserialized wrapper.

fn model_type() -> ModelType

Get the model type identifier.

impl ModelSave for RegressionOutput

fn save_with_name(&self, path: &str, name: Option<String>) -> Result<()>

Save the model to a file with a custom name.

fn model_type() -> ModelType

Get the model type identifier.

fn save(&self, path: &str) -> Result<(), Error>

Save the model to a file.

impl Serialize for RegressionOutput

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

where __S: Serializer,

Serialize this value into the given Serde serializer.