Skip to main content

Encoder

axon_encoder

Trait Encoder 

Source 
pub trait Encoder {
    // Required methods
    fn encode(&mut self, input: &[f32]) -> EncodedOutput;
    fn reset(&mut self);

    // Provided method
    fn encode_step(&mut self, input: &[f32]) -> EncodedOutput { ... }
}
Expand description

The core trait for all encoders in this crate.

Encoders convert continuous analog values into discrete spike events for spiking neural networks (SNNs). Two modes are supported:

  • Batch mode (encode): Process a complete input vector at once.
  • Streaming mode (encode_step): Process incrementally, one step at a time.

§Example

use axon_encoder::prelude::*;

let mut encoder = RateEncoder::try_new(5.0, 50.0, (0.0, 1.0), 0.010)?;
let input = [0.25, 0.75, 0.5];

// Batch encoding
let output = encoder.encode(&input);

// Reset for streaming (if using stateful encoder)
encoder.reset();

Required Methods§

Source

fn encode(&mut self, input: &[f32]) -> EncodedOutput

Encodes a slice of analog values into spike events (batch mode).

Source

fn reset(&mut self)

Resets the encoder to its initial state

Provided Methods§

Source

fn encode_step(&mut self, input: &[f32]) -> EncodedOutput

Encodes a single step incrementally (streaming mode).

By default, this delegates to encode() for stateless encoders. Stateful encoders should override this to maintain state between calls.

§Arguments
  • input - A slice of analog values to encode
§Returns

An EncodedOutput containing any spike events generated in this step

Dyn Compatibility§

This trait is dyn compatible.

In older versions of Rust, dyn compatibility was called "object safety".

Implementors§