Spacechop
Open
- Dominant language
- Jupyter Notebook
- Stars
- 201
- Forks
- 62
- PR merge metrics
- No merged PRs in 30d
Description
Geographic cross validation: train on geographic unit A and test on geographic unit B. Can help choose models that perform well across space, identify new spatial features, and estimate accuracy of model when applied to new geographies.
After spacechop, can do spacetimechop: train on geographic unit A in time t and test on geographic unit B in time t+1
Contributor guide
Assessment
This issue has not been assessed yet.