Struct TemporalEncoder 

pub struct TemporalEncoder { /* private fields */ }

Encodes temporal patterns by tracking history of values per channel.

Fires a spike when the rate of change exceeds configurable thresholds. Useful for detecting sudden changes or motion in sensor signals.

Mathematical Model

Computes the difference between recent average (last 3 values) and older average (previous 3 values before that). A spike is generated when this change exceeds the threshold:

change = |mean(history[-3:]) - mean(history[-6:-3])|
spike if change > threshold

When to Use

• Detecting sudden changes in signal (edge detection)
• Motion detection in video or sensor streams
• Event-based encoding where changes are more important than absolute values

Parameters

• history_depth: How many past values to track per channel
• change_thresholds: Vec of (threshold, spike_value) pairs - fires when change exceeds threshold
• num_channels: Number of input channels

Examples

use axon_encoder::prelude::*;
// history_depth must be at least 6 for the dual-window change detector.
let mut enc = TemporalEncoder::try_new(6, vec![(0.5, 1)], 1)?;
for v in [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 2.0] {
    let _ = enc.encode_step(&[v]);
}

Implementations

impl TemporalEncoder

pub fn new( history_depth: usize, change_thresholds: Vec<(f32, u16)>, num_channels: usize, ) -> Self

Creates a new TemporalEncoder, panicking if configuration is invalid.

Prefer try_new for typed validation errors.

Panics

Panics if history_depth < 6 or num_channels is unsupported.

pub fn try_new( history_depth: usize, change_thresholds: Vec<(f32, u16)>, num_channels: usize, ) -> Result<Self, EncoderError>

Creates a new TemporalEncoder, returning an EncoderError for invalid configuration.

Each threshold in change_thresholds must be finite and non-negative.

pub fn encode_with_modulators( &mut self, input: &[f32], modulators: &NeuroModulators, gain_curves: &NeuromodulatorGainCurves, ) -> EncodedOutput

Encodes input using neuromodulator-driven gain curves.

Inherent wrapper so callers need not import ModulatedEncoder.

pub fn encode_step_with_modulators( &mut self, input: &[f32], modulators: &NeuroModulators, gain_curves: &NeuromodulatorGainCurves, ) -> EncodedOutput

Step-wise variant of encode_with_modulators.

Trait Implementations

impl Clone for TemporalEncoder

fn clone(&self) -> TemporalEncoder

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for TemporalEncoder

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

Formats the value using the given formatter.

impl Encoder for TemporalEncoder

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

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

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

Encodes a single step incrementally (streaming mode).

fn reset(&mut self)

Resets the encoder to its initial state

impl ModulatedEncoder for TemporalEncoder

fn encode_with_gains( &mut self, input: &[f32], gains: EncodingGains, ) -> EncodedOutput

Encodes input using already evaluated encoding gains.

fn encode_step_with_gains( &mut self, input: &[f32], gains: EncodingGains, ) -> EncodedOutput

Encodes one streaming step using already evaluated encoding gains.

fn encode_with_modulators( &mut self, input: &[f32], modulators: &NeuroModulators, gain_curves: &NeuromodulatorGainCurves, ) -> EncodedOutput

Encodes input using neuromodulator-driven gain curves.

fn encode_step_with_modulators( &mut self, input: &[f32], modulators: &NeuroModulators, gain_curves: &NeuromodulatorGainCurves, ) -> EncodedOutput

Encodes one streaming step using neuromodulator-driven gain curves.

impl PartialEq for TemporalEncoder

fn eq(&self, other: &TemporalEncoder) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl StructuralPartialEq for TemporalEncoder