An open scoring tool for crop-modelling datasets
How good
is your dataset?
DataRanking applies the Kersebaum et al. 2015 framework to evaluate agricultural field datasets across eight blocks — management, phenology, previous crop, initial values, soil, site, weather, and state variables — and assigns each block and the dataset as a whole one of four tiers: Platinum, Gold, Silver, or Copper.
- 01
Paste a dataset
Drop a JSON file, paste it in, or load the Münchenberg example to see the framework in motion. Same shape the original 2014 desktop tool used — one nested object keyed by block.
- 02
Score it
The pure-Python port reproduces the original C++
RankPointGeneratorto the byte. Every block formula, the seasons multiplier, and the tier thresholds match the shipped binary. - 03
Read the verdict
One headline tier, eight per-block tiers, a continuous score and an instrument-readable radar of where the dataset is strong and where it’s thin.
Known-good
Müncheberg (1992–1998)
Christian Kersebaum’s own ranking of the Müncheberg crop-rotation dataset — shipped with the 2014 Windows distribution — was the first thing we pinned the port against. The web app returns the same verdict the 2014 binary did:
| Raw | × mult. | = Adj. | Verdict |
|---|---|---|---|
| 162.64 | 1.0857 | 176.58 | Gold |
The fixture and its expected verdict live in the repo as
a pytest regression suite — 21 assertions
across all eight blocks, the multiplier, the raw and adjusted
totals, and the overall tier.