How shekhar-commit/fertilizer_recommendation_system works
This is a Flask fertilizer-recommendation web app. A user submits crop, soil, and environmental inputs to the prediction route; the app converts numeric fields, encodes categorical features, predicts a fertilizer, applies quantity/timing/method rules, and renders a recommendation. The model and encoders are loaded from repository artifacts. The notebook and styling are not shown as runtime components. Source excerpts support the app’s calls and template references; template contents were not sampled, so their internal behavior is not inferred.
Main entry points: app.py Prediction route · result.html Result page · encoders.pkl Feature encoders · index.html Input page
Components 11
Web interface
- Home route
app.py - Input page
templates/index.html - Prediction route
app.py - Result page
templates/result.html
Prediction
- Numeric conversion
app.py - Feature encoding
app.py - Feature encoders
encoders.pkl - Fertilizer model
fertilizer_model.pkl - Fertilizer decoding
app.py
Recommendation
- Recommendation rules
app.py
Other components
- Farmer
Connections 11
- Farmer → Prediction route: submits inputs
app.py - Home route → Input page: renders
app.py - Prediction route → Numeric conversion: converts numbers
app.py - Prediction route → Recommendation rules: requests recommendation
app.py - Recommendation rules → Feature encoding: predicts fertilizer
app.py - Feature encoding → Feature encoders: transforms features
app.py - Feature encoding → Fertilizer model: predicts class
app.py - Feature encoding → Fertilizer decoding: decodes prediction
app.py - Fertilizer decoding → Feature encoders: inverse transforms
app.py - Recommendation rules → Result page: renders recommendation
app.py - Prediction route → Result page: renders result
app.py