Releases: SantanderMetGroup/downscaleR
Releases · SantanderMetGroup/downscaleR
v3.0.5
New in v3.0.5 (17 Dic 2018)
- Add important explanation in the biasCorrection documentation
- Bug fix in argument which.combine of prepareData
- Adapt biasCorrection to grids with a single dimension (time)
- NA handling in precipitation frequency adjustment in biasCorrection
- Code improvement
climate4R CITATION
v3.0.4 (23 Sep 2018)
- Add new CITATION
- Edited startup package message
v3.0.3
v3.0.2
v3.0.1
v3.0.0
New in downscaleR v3
- New user interface for flexible definition of predictors (
prepareData
) and prediction data (prepareNewData
). - New method (neural networks) available for perfect-prog downscaling
- New workhorse function
downscale
for perfect-prog/MOS downscaling with different methods (GLM, Analogs..., more soon) - Flexible method calibration/prediction via
downscale.train
anddownscale.predict
- New helper function for flexible cross-validation experimental setups:
downscale.cv
- Improved efficiency in
biasCorrection
, including a parallelization option - New options for a more flexible parametric quantile mapping method design
- All changes in predictor/predictand definition have been double-checked against MeteoLab for consistency
- Other documentation updates and enhancements
- See the updates in the wiki for worked examples