Paper-Aligned Toolbox Map

This page aligns the nSTAT toolbox documentation with the original toolbox paper:

Full text links:

Contents

Class Hierarchy and Object Model

nSTAT is organized around reusable signal and trial abstractions.

Class references and examples:

Fitting and Assessment Workflow

The paper's core workflow fits point-process GLMs and evaluates fit quality.

1. Build trial data with `Trial`, `CovColl`, and `nstColl`. 2. Define candidate models with `TrialConfig` and `ConfigColl`. 3. Fit models with `Analysis.RunAnalysisForNeuron` or `Analysis.RunAnalysisForAllNeurons`. 4. Assess goodness-of-fit using `FitResult` diagnostics (KS, residuals, confidence bands) and summarize across neurons with `FitResSummary`.

Related examples:

Simulation Workflow

The toolbox supports simulation of point-process and related neural models.

Decoding Workflow

nSTAT includes point-process and Gaussian-state decoding algorithms that are described in the paper's adaptive filtering sections.

Example-to-Paper Section Mapping

The examples below correspond directly to the paper's representative workflows.