Data Mining Algorithms In R/Packages/CCMtools/learn.and.project.clusters
Jump to navigation
Jump to search
Description[edit | edit source]
This function (1) learns how to attribute days to clusters based on the sequence of predictors and associated sequence of clusters.
Usage[edit | edit source]
learn.and.project.clusters(DataCalibration, DataToBeProjected, cl.calibration, allocmet, Datas.Calibration)
Arguments[edit | edit source]
- DataCalibration Values of the predictor variable for the calibration set (can be a matrix).
- DataToBeProjected Values of the predictor variable for the projection set.
- cl.calibration Numerical vector corresponding to the sequence of clusters (i.e., calibration set).
- allocmet Name of the attribution method.(The 12 possibilities are: "Euclid.dist.A", "Euclid. dist.w1", "Euclid.dist.w2", "CART.A", "CART.w", "CART.A.and.w", "knnA", "knnA10", "Gaussian.A", "Gaussian.w", "MM", "MMw")
- DataS.Calibration Values of other predictor variables for the calibration set. This is sometimes needed, according to the attribution method (allocmet) to be used (needed for "Euclid.dist.w1", "Euclid.dist.w2", "CART.w", "CART.A.and.w", "Gaussian.w", "MMw").
Value[edit | edit source]
Returns a list with two objects:
- cl The sequence of clusters defined from the predictors for the projection set.
- tot Number of elements per cluster for projection set.
Author[edit | edit source]
M. Vrac (mathieu.vrac@lsce.ipsl.fr))