pasture
OpenML dataset with id 339
No author found.
Full work available at URL: https://api.openml.org/data/v1/download/52242/pasture.arff
Upload date: 26 August 2014
Dataset Characteristics
Number of classes: 3
Number of features: 23 (numeric: 21, symbolic: 2 and in total binary: 0 )
Number of instances: 36
Number of instances with missing values: 0
Number of missing values: 0
Author: Dave Barker Source: original - Please cite:
Pasture Production
Data source: Dave Barker AgResearch Grasslands, Palmerston North, New Zealand
The objective was to predict pasture production from a variety of biophysical factors. Vegetation and soil variables from areas of grazed North Island hill country with different management (fertilizer application/stocking rate) histories (1973-1994) were measured and subdivided into 36 paddocks. Nineteen vegetation (including herbage production); soil chemical, physical and biological; and soil water variables were selected as potentially useful biophysical indicators.
Attribute Information:
1. fertiliser - fertiliser used - enumerated
2. slope - slope of the paddock - integer
3. aspect-dev-NW - the deviation from the north-west - integer
4. OlsenP - integer
5. MinN - integer
6. TS - integer
7. Ca-Mg - calcium magnesium ration - real
8. LOM - soil lom (g/100g) - real
9. NFIX-mean - a mean calculation - real
10. Eworms-main-3 - main 3 spp earth worms per g/m2 - real
11. Eworms-No-species - number of spp - integer
12. KUnSat - mm/hr - real
13. OM - real
14. Air-Perm - real
15. Porosity - real
16. HFRG-pct-mean - mean percent - real
17. legume-yield - kgDM/ha - real
18. OSPP-pct-mean - mean percent - real
19. Jan-Mar-mean-TDR - real
20. Annual-Mean-Runoff - in mm - real
21. root-surface-area - m2/m3 - real
22. Leaf-P - ppm - real
23. pasture-prod-class - pasture production categorisation - enumerated
