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.

Try it → Read the algorithm See the gallery

  1. 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.

  2. 02

    Score it

    The pure-Python port reproduces the original C++ RankPointGenerator to the byte. Every block formula, the seasons multiplier, and the tier thresholds match the shipped binary.

  3. 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.

Provenance

The framework is published in:

Kersebaum, K.C., K.J. Boote, J.S. Jorgenson, C. Nendel, M. Bindi, C. Frühauf, T. Gaiser, G. Hoogenboom, C. Kollas, J.E. Olesen, R.P. Rötter, F. Ruget, P.J. Thorburn, M. Trnka, M. Wegehenkel (2015). Analysis and classification of data sets for calibration and validation of agro-ecosystem models. Environmental Modelling & Software 72: 402–417. DOI.