Trait rusty_machine::learning::k_means::Initializer

pub trait Initializer: Debug {
    fn init_centroids(&self,
                  k: usize,
                  inputs: &Matrix<f64>)
                  -> LearningResult<Matrix<f64>>;
}

Trait for algorithms initializing the K-means centroids.

Required Methods

fn init_centroids(&self,
                  k: usize,
                  inputs: &Matrix<f64>)
                  -> LearningResult<Matrix<f64>>

Initialize the centroids for the initial state of the K-Means model.

The Matrix returned must have k rows and the same column count as inputs.

Implementors

• impl Initializer for Forgy
• impl Initializer for RandomPartition
• impl Initializer for KPlusPlus