Cassava fertilizer decision support for Nigeria
# Cassava Fertilizer Advisor -- Nigeria
A simple, non-technical dashboard for extension agents and advisors,
translating the climate-trend-adjusted, QUEFTS-based nutrient model into a
plain-language fertilizer recommendation.
## What it does
- Pick a Nigerian state from a dropdown
- Tell it whether your soil allows deep rooting (or "I'm not sure")
- Get a recommendation in bags of Urea, TSP, and MOP per hectare (the
fertilizer products actually sold at Nigerian agro-dealers), not raw kg/ha
of elemental nutrients
- See it on a real state-boundary map
- Expand "How trustworthy is this?" for the honest validation story in plain
language -- including that the first version was wrong and was corrected
## How to run it
```bash
pip install streamlit plotly pandas
streamlit run app.py
```
Then open the local URL it prints (usually `
localhost`).
## Files needed together in the same folder
- `app.py` -- the dashboard itself
- `state_summary.csv` -- state-level fertilizer recommendations, pre-computed
from the full analysis (see the manuscript / Colab notebook for how this
was derived)
- `nigeria_states.geojson` -- real Nigerian state boundaries, for the map
- `requirements.txt` -- tells Streamlit Cloud which Python packages to install
## Deploying to Streamlit Community Cloud (free, public link)
This needs your own GitHub account and your own Streamlit Community Cloud
account -- I can't create either on your behalf, but every step below is
exact and copy-paste ready.
**Step 1 -- Create a GitHub repository**
1. Go to github.com and sign in (or create a free account if you don't have one).
2. Click the "+" in the top-right corner, then "New repository."
3. Name it something like `cassava-fertilizer-advisor`. Set it to **Public**
(Community Cloud's free tier deploys public repos most simply; private
repos work too but need an extra permission step later).
4. Click "Create repository."
**Step 2 -- Upload the four files**
1. On your new repository's page, c …