Struct RetinalGanglionCell 

pub struct RetinalGanglionCell {
    pub stim_buffer: Vec<f64>,
    pub stim_kernel: Vec<f64>,
    pub stim_idx: usize,
    pub hist_buffer: Vec<f64>,
    pub hist_kernel: Vec<f64>,
    pub hist_idx: usize,
    pub baseline: f64,
    pub on_centre: bool,
    pub spike_threshold: f64,
    pub dt: f64,
    pub gain: f64,
}

Retinal ganglion cell — Pillow et al. 2005 GLM.

Generalized linear model (GLM) for retinal ganglion cells, the gold standard for statistical spike train models:

1. Stimulus filter (k): temporal kernel convolved with stimulus. Implemented as a causal FIR filter over a ring buffer of past stimulus values. Default: biphasic filter (fast excitatory + slow inhibitory lobe), ON-centre or OFF-centre.

2. Post-spike history filter (h): self-feedback after each spike. Models absolute/relative refractoriness and burst facilitation. Implemented as exponential basis functions applied to spike history. Default: strong inhibitory (refractory) followed by weak excitatory (burst tendency).

3. Exponential nonlinearity: λ(t) = exp(k·s(t) + h·spike_history + b) where λ is the instantaneous firing rate (Hz).

4. Spike generation: deterministic threshold on λ(t). Spike emitted when λ(t) * dt > threshold (proxy for inhomogeneous Poisson at high rate).

Pillow et al., Nature 437:1258, 2005. Pillow et al., J Neurosci 28:11003, 2008 (coupled GLM).

State: stimulus ring buffer, spike history ring buffer, filtered stimulus value, filtered history value.

Fields

stim_buffer: Vec<f64>

stim_kernel: Vec<f64>

stim_idx: usize

hist_buffer: Vec<f64>

hist_kernel: Vec<f64>

hist_idx: usize

baseline: f64

on_centre: bool

spike_threshold: f64

dt: f64

gain: f64

Implementations

impl RetinalGanglionCell

pub fn new() -> Self

Create ON-centre RGC with default biphasic stimulus filter and post-spike history filter.

pub fn off_centre() -> Self

fn convolve(buffer: &[f64], kernel: &[f64], write_idx: usize) -> f64

Convolve ring buffer with kernel (dot product with circular indexing).

pub fn step(&mut self, input: f64) -> i32

Step with bipolar cell input. Returns spike (1/0).

GLM pipeline: stimulus filter → history filter → exp nonlinearity → spike

pub fn reset(&mut self)

Trait Implementations

impl Clone for RetinalGanglionCell

fn clone(&self) -> RetinalGanglionCell

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for RetinalGanglionCell

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

Formats the value using the given formatter.

impl Default for RetinalGanglionCell

fn default() -> Self

Returns the “default value” for a type.