NEWS.md
IKARUS 0.1.0.9000
development version for LLOCV logic
-
new features extract_tDat_LLOCV - equal to 'extract_tDat' but uses a "location" column and saves the location for each pixel to returning dataframe RFclass_LLOCV - equal to 'RFclass' but uses LLOCV instead of simple random CV, requires a tDat with class and location information. nk is a parameter to select the k folds for CV. BestPredFFS_LLOCV - equal to 'BestPredFFS' but uses LLOCV instead of simple random CV, requires a tDat with class and location and comnined class_location information. nk is a parameter to select the k folds for CV.
-
add example data lau_TrainPoly_LLOCV2 - contains location informations, has smaller polygons than 'lau_TrainPoly'
IKARUS 0.1.0
release version
- add IKARUS_usage.html
IKARUS 0.0.0.93
-
add vignette
-
some bugfixing ClassSegVal - now example calls correct function name ClassSegVal - help now contains description for the resulting tabel ClassSegVal - some changes in code and returning table RFclass - now example calls correct variable
IKARUS 0.0.0.92
-
major spelling corrections
-
simplified example data loading
IKARUS 0.0.0.91
*new feature classSegVal - validation of Tree classification by Tree Crown Segments
IKARUS 0.0.0.9000
initial version
*new feature BestPredFFS - Forward feature selection for training data. extract_Traindat - function for extracting Training Data from a Raster Object. RFclass - classification function for RandomForest. IKARUS - Package - Package informations and Example data documentation.
