trspecfit — 2D Time- and Energy-resolved Spectroscopy Fitting

A Python library for fitting multi-component spectral models to time-resolved spectroscopy data.

User Guide:

• Installation
  • From PyPI
  • For Included Example Notebooks
  • From GitHub
  • For Development
  • Step-by-step for macOS and Linux
  • Step-by-step for Windows
• Quick Start
  • Get running
  • Examples overview
  • Typical workflow
  • YAML model format
  • Model composition rules
• Examples
  • Fitting Workflows
  • Data Generation
  • Notebooks
• Supported Models
  • Supported Today
  • Expression Support Today
  • Explicitly Excluded Today
  • Lowering Notes
  • Notes

Developer Info:

• Repo Architecture
  • Two-layer design
  • Top-level modules (src/trspecfit/)
  • Fit results: save / export / load architecture
  • config/ — runtime configuration
  • functions/ — the function registry
  • utils/ — reusable helpers (check here before inventing)
  • Typical execution flow
  • Where to put new code — quick guide
• Execution Plan: Lowered Evaluator
  • Motivation
  • Architecture overview
  • Spec
  • Implementation tracking
• Fit-archive HDF5 schema (schema_version 2)
  • Conventions
  • Top-level layout
  • File group
  • Slot group
  • params dataset
  • metrics_per_slice dataset (sbs only)
  • conf_ci dataset (optional)
  • mcmc/ group (optional)
  • Reader → object-model mapping
  • Per-fit-type cheat sheet
  • What’s not in v1
  • Cross-references
• Roundtrip Test Matrix
  • Axes
  • Canonical model families
  • Target matrix
  • Project-level matrix
  • MCMC and worker policy
  • Practical rule
  • Minimum test shape per cell
  • Current snapshot
  • Suggested implementation order
  • Non-goals
• Examples Upgrade Plan
  • Decisions (locked)
  • Motivation
  • Current State
  • User Workflows
  • Proposed Directory Layout
  • Notebook Content Targets
  • Docs Navigation
  • Save/Export/Load Presentation
  • Migration Plan
  • Non-goals For The Save/Load Branch
  • Open Questions
• Skills
  • Add Function
  • Benchmark GIR vs Interpreter
  • Changelog Workflow
  • Check Docs
  • Code Review Checklist

Reference:

• API Reference
  • Main Module
  • Model/Component/Parameter Module
  • Energy Functions
  • Time Functions
  • Profile Functions
  • Simulator Module
  • Plot Configuration
  • Fitting Module
  • Spectra Module
  • GraphIR and Scheduler
  • 1D Evaluator
  • 2D Evaluator

Overview

trspecfit is a Python library to define and fit multi-component spectral models for 1D energy-resolved and 2D time-and-energy-resolved spectroscopy data. The package extends lmfit-based workflows to support component-based models, time-dynamics, convolution kernels and helpers for generating simulated data.

Key Capabilities

• Define modular spectral components (Gaussian, Voigt/GLP/GLS, Doniach–Sunjic, backgrounds, convolution kernels)

• Build 1D (energy) and 2D (time × energy) spectra from components with time-dynamics

• Attach auxiliary-axis parameter profiles (e.g. depth/position) and compose with dynamics

• Fit models with lmfit and wrappers for confidence intervals and optional MCMC sampling

• Simulate individual 1D/2D spectra to validate models

• Generate ML training data via parameter space exploration

Design Guidance

For the current supported model-composition rules and expression semantics, see Supported Models. This is the reference guide for which model combinations are intended to work today, which are explicitly excluded, and which edge cases are tolerated for now but may change in the future.

Indices and tables

• Index

• Module Index

• Search Page