API Reference

This section provides detailed documentation for all modules, classes, and functions in OpenRetina.

Core Components

• Models: Pre-trained and customizable retina models
• Modules: Building blocks for constructing models
• Data I/O: Data loading and processing utilities
• In-silico: Tools for analyzing models with virtual experiments
• Utilities: Helper functions and visualization tools

API Documentation by Module

The OpenRetina package is organized in a modular structure where each module serves a specific purpose in the retina modeling workflow.

Models

The models module provides complete model implementations:

• openretina.models.core_readout: End-to-end convolutional models with spatial readouts
• openretina.models.linear_nonlinear: Classical linear-nonlinear cascade models
• openretina.models.sparse_autoencoder: Models for learning sparse representations

Modules

The modules package contains building blocks for model construction. See the modules overview for complete documentation.

Data I/O

The data_io package handles data loading and preprocessing:

• openretina.data_io.base_dataloader: Abstract interfaces for data handling
• openretina.data_io.hoefling_2024: Dataloaders for Höfling et al. 2024 dataset
• openretina.data_io.artificial_stimuli: Utilities for generating artificial stimuli
• openretina.data_io.cyclers: Utilities for cycling through datasets

See the data I/O overview for complete documentation.

In-silico

The insilico package provides tools for analyzing models:

• openretina.insilico.stimulus_optimization: Tools for finding optimal stimuli

See the in-silico overview for complete documentation.