API Reference

Complete API documentation for NLSQ modules and functions.

• NLSQ API Reference
  • Core API
  • Pre-Built Functions
  • Large Dataset Support
  • Adaptive Hybrid Streaming (v0.3.0+)
  • Global Optimization (v0.3.3+)
  • Core Factories (v0.4.3+)
  • Orchestration Components (v0.6.4+)
  • Facades (v0.6.4+)
  • Workflow System (v0.5.5+)
  • Command-Line Interface (v0.4.1+)
  • Qt Desktop GUI (v0.5.0+)
  • Enhanced Features (v0.1.1)
  • Interfaces & Protocols (v0.4.2+)
  • Algorithms & Optimization
  • Utilities & Infrastructure
  • Performance & Profiling
  • Module Index
  • Complete Module Listing
• Notebook Utilities API Reference
  • Module Overview
  • Type Definitions
  • Cell Utilities
  • Core I/O Operations
  • Transformation Base Class
  • Transformation Implementations
  • Pipeline Orchestration
  • Incremental Processing
  • See Also

Core API

The main NLSQ API provides drop-in replacements for SciPy’s curve fitting functions:

Main Functions

• nlsq.fit() - Unified curve fitting with preset-based configuration (recommended)

• nlsq.curve_fit() - High-level curve fitting interface (SciPy-compatible)

• nlsq.curve_fit_large() - Automatic chunking for large datasets

• nlsq.LeastSquares - Low-level least squares solver class

• nlsq.CurveFit - Reusable curve fitting class (JIT-compiled)

See NLSQ API Reference for complete module documentation.

Large Dataset API

Specialized functions for large-scale fitting:

• nlsq.curve_fit_large() - Automatic chunking and memory management

• nlsq.LargeDatasetFitter - Advanced dataset management

See Large Dataset API Reference for detailed documentation.

Adaptive Hybrid Streaming API

Four-phase hybrid optimizer combining parameter normalization, L-BFGS warmup, streaming Gauss-Newton, and exact covariance computation:

• nlsq.AdaptiveHybridStreamingOptimizer - Main optimizer class

• nlsq.HybridStreamingConfig - Configuration with presets

• nlsq.ParameterNormalizer - Parameter normalization

See:

• nlsq.adaptive_hybrid_streaming module - Main optimizer documentation

• nlsq.hybrid_streaming_config module - Configuration options

• nlsq.parameter_normalizer module - Parameter normalization

Module Organization

Core Modules

• nlsq.minpack - Main curve_fit implementation

• nlsq.least_squares - Least squares solver

• nlsq.trf - Trust Region Reflective algorithm

Advanced Features

• nlsq.large_dataset - Large dataset handling

• nlsq.memory_manager - Memory management

• nlsq.smart_cache - Intelligent caching

• nlsq.diagnostics - Model Health Diagnostics System (identifiability, gradient health, sloppy model analysis)

• nlsq.adaptive_hybrid_streaming - Four-phase hybrid optimizer

• nlsq.hybrid_streaming_config - Hybrid streaming configuration

• nlsq.parameter_normalizer - Parameter normalization

Utilities

• nlsq.validators - Input validation

• nlsq.loss_functions - Loss function library

• nlsq.config - Configuration management

• nlsq.logging - Logging utilities

See NLSQ API Reference for complete documentation of all modules.

Development Tools

Notebook Configuration Utilities

Modern framework for transforming Jupyter notebooks with automated configurations:

• notebook_utils - Notebook transformation package

• TransformationPipeline - Pipeline orchestration

• ProcessingTracker - Incremental processing

See Notebook Utilities API Reference for complete API documentation and Notebook Configuration Utilities for usage guide.

Performance Benchmarks

See NLSQ Optimization Case Study: When to Stop Optimizing for detailed performance analysis.