Practical Guides

Welcome to the practical guides for the AstroSpectro project. This section walks you through each key step of the pipeline with concrete, task-oriented tutorials.

The goal here is to show you how to complete specific actions: use a pipeline step independently, explore a function in greater depth, or adapt parameters to your needs.

Start here

New to AstroSpectro? Follow this route:

1. Installation: prepare your Python environment, or open a Codespace.
2. Quick Start: run the whole pipeline once, from downloading spectra to the first predictions.
3. Then go deeper: the guides below for each step, Concepts & Architecture for the reasoning, the API Reference for classes and functions.

Stuck? Check the FAQ or open an issue.

Available Guides

Each guide focuses on one part of the workflow. You do not have to follow them in order.

Related page	Relation
Data download	Retrieve thousands of spectra from LAMOST DR5 and manage the local catalogue
Spectrum preprocessing	Normalise and clean spectra so the models can use them
Feature extraction	Detect spectral lines and turn them into numerical features
Model training	Train a supervised classifier on the extracted features
Results visualisation	Inspect spectra, detected lines and model performance

Going deeper

To understand the theory and the "why" behind each step, see Concepts & Architecture.