sc_neurocore_engine/bindings/stochastic/
glm.rs

1// SPDX-License-Identifier: AGPL-3.0-or-later
2// Commercial license available
3// © Concepts 1996–2026 Miroslav Šotek. All rights reserved.
4// © Code 2020–2026 Miroslav Šotek. All rights reserved.
5// ORCID: 0009-0009-3560-0851
6// Contact: www.anulum.li | protoscience@anulum.li
7// SC-NeuroCore — Generalised linear model neuron PyO3 binding
8
9use pyo3::exceptions::PyValueError;
10use pyo3::prelude::*;
11use pyo3::types::PyDict;
12
13use crate::neurons;
14
15#[pyclass(name = "GLMNeuron", module = "sc_neurocore_engine.sc_neurocore_engine")]
16#[derive(Clone)]
17pub struct PyGLMNeuron {
18    inner: neurons::GLMNeuron,
19}
20
21#[pymethods]
22impl PyGLMNeuron {
23    #[new]
24    #[pyo3(signature = (n_k=10, n_h=20, seed=42))]
25    fn new(n_k: usize, n_h: usize, seed: u64) -> Self {
26        Self {
27            inner: neurons::GLMNeuron::new(n_k, n_h, seed),
28        }
29    }
30
31    /// Construct the original engine constant-filter configuration without
32    /// adding a second model identity to the public catalogue.
33    #[staticmethod]
34    #[pyo3(signature = (n_k=10, n_h=20, seed=42))]
35    fn legacy_constant_filters(n_k: usize, n_h: usize, seed: u64) -> Self {
36        Self {
37            inner: neurons::GLMNeuron::new_legacy_constant_filters(n_k, n_h, seed),
38        }
39    }
40
41    /// Advance one bin; `uniform` (in `[0, 1)`) makes the Bernoulli draw
42    /// deterministic, otherwise the seeded generator is used. Raises
43    /// `ValueError` with the history unchanged on any invalid input.
44    #[pyo3(signature = (stimulus, uniform=None))]
45    fn step(&mut self, stimulus: f64, uniform: Option<f64>) -> PyResult<i32> {
46        self.inner
47            .try_step(stimulus, uniform)
48            .map_err(PyValueError::new_err)
49    }
50
51    /// Clear both history buffers, preserving the filters and generator.
52    fn reset(&mut self) {
53        self.inner.reset();
54    }
55
56    /// Return filters, parameters, and history buffers as a Python dictionary.
57    fn get_state(&self, py: Python<'_>) -> PyResult<Py<PyAny>> {
58        let state = PyDict::new(py);
59        state.set_item("mu", self.inner.mu)?;
60        state.set_item("dt_ms", self.inner.dt_ms)?;
61        state.set_item("k", self.inner.k.clone())?;
62        state.set_item("h", self.inner.h.clone())?;
63        state.set_item("stim_buf", self.inner.stim_buf_view())?;
64        state.set_item("spike_buf", self.inner.spike_buf_view())?;
65        Ok(state.into_any().unbind())
66    }
67}
68
69pub(super) fn register(module: &Bound<'_, PyModule>) -> PyResult<()> {
70    module.add_class::<PyGLMNeuron>()?;
71    Ok(())
72}
sc_neurocore_engine/bindings/stochastic/glm.rs

sc_neurocore_engine/bindings/stochastic/
glm.rs
