How radiah23/stork-tracking---sequential-data-analysis works
This repository tracks white-stork migration from raw GPS fixes through cleaning, sequencing, feature extraction, SQLite storage, SQL analysis, and mapping; it also offers an LLM question-answering path over the stored data. The README documents the pipeline and analytical capabilities. The visualization implementation is not present in the tree, so the map is represented only by the available map.html artifact; links involving it reflect documented behavior, not inspected code.
Main entry points: ask_data.py Question Workflow · extract.py Tracking Database · run_queries.py Query Runner · queries.sql SQL Analyses
Components 13
GPS Preparation
- CSV Extraction
extract.py - Quality Cleaning
extract.py - Sequence Ordering
extract.py - Movement Features
extract.py - Tracking Database
extract.py
Analysis and Mapping
- SQL Analyses
queries.sql - Query Runner
run_queries.py - Migration Map
map.html
Natural Language Queries
- Question Workflow
ask_data.py - SQL Safety
ask_data.py
Other components
- Movebank GPS CSV
- Data User
- Anthropic Claude
Connections 13
- Movebank GPS CSV → CSV Extraction: provides CSV
README.md - CSV Extraction → Quality Cleaning: loads rows
README.md - Quality Cleaning → Sequence Ordering: orders fixes
README.md - Sequence Ordering → Movement Features: derives features
README.md - Movement Features → Tracking Database: writes table
extract.py - Query Runner → SQL Analyses: loads SQL
run_queries.py - Query Runner → Tracking Database: queries
run_queries.py - SQL Analyses → Migration Map: supports visualization
README.md - Data User → Question Workflow: asks question
README.md - Question Workflow → Anthropic Claude: requests SQL
ask_data.py - Question Workflow → SQL Safety: validates query
ask_data.py - SQL Safety → Tracking Database: reads only
ask_data.py - Question Workflow → Anthropic Claude: requests explanation
ask_data.py