Struct DistributedLinearRegression 

pub struct DistributedLinearRegression {
    pub config: DistributedConfig,
    pub parameter_server: Arc<RwLock<ParameterServer>>,
    pub workers: Vec<WorkerNode>,
    pub parameters: Arc<RwLock<ModelParameters>>,
}

Distributed linear regression using parameter server architecture

Implements distributed training of linear regression models across multiple worker nodes with centralized parameter synchronization.

Fields

config: DistributedConfig

Configuration for distributed training

parameter_server: Arc<RwLock<ParameterServer>>

Parameter server for coordinating updates

workers: Vec<WorkerNode>

Worker nodes performing computation

parameters: Arc<RwLock<ModelParameters>>

Current model parameters

Implementations

impl DistributedLinearRegression

pub fn new(config: DistributedConfig, num_features: usize) -> Self

Create a new distributed linear regression model

pub fn partition_data( &mut self, features: Vec<Vec<f64>>, targets: Vec<f64>, ) -> Result<()>

Partition data across workers

pub fn fit(&mut self) -> Result<()>

Train the model using distributed gradient descent

pub fn get_training_stats(&self) -> DistributedTrainingStats

Get training statistics from all workers

pub fn predict(&self, features: &[Vec<f64>]) -> Result<Vec<f64>>

Predict on new data using trained parameters

Trait Implementations

impl Debug for DistributedLinearRegression

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

Formats the value using the given formatter.