How vishaldharsan27-ux/marine_ai works
This repository currently exposes a command-line fisheries modeling workflow: running main.py loads fisheries and weather data, cleans fisheries records, reshapes them for modeling, then trains and evaluates a catch model. The README describes preprocessing and prediction capabilities, but the corresponding src/ml/ files are absent from the supplied tree; their nodes therefore use null paths. The weather data directory is also only a documented placeholder, and no user-facing prediction service is implemented.
Main entry points: main.py Pipeline entry · Table_A-5.5.csv Catch statistics
Components 9
Data inputs
- Catch statistics
data/raw/cmfri/Table_A-5.5.csv - Weather data
Modeling pipeline
- Pipeline entry
main.py - Data loading
- Data cleaning
- Feature engineering
Catch modeling
- Catch model
- Model evaluation
main.py
Other components
- Pipeline operator
Connections 9
- Pipeline operator → Pipeline entry: runs
README.md - Pipeline entry → Data loading: loads data
main.py - Catch statistics → Data loading: supplies statistics
README.md - Weather data → Data loading: supplies weather
README.md - Pipeline entry → Data cleaning: cleans fisheries
main.py - Pipeline entry → Feature engineering: engineers features
main.py - Pipeline entry → Catch model: trains model
main.py - Pipeline entry → Model evaluation: evaluates model
main.py - Catch model → Model evaluation: provides trained model
main.py