Last chance! 50% off unlimited learning
Sale ends in
dudiLeave-one-out cross-validation to check the dispersion of row coordinates in a dudi.
# S3 method for dudi
loocv(x, progress = FALSE, ...)A list with:- XValCoord:
the cross-validated row coordinates - PRESS:
the Predicted Residual Error Sum for each row- PRESSTot:
the sum of PRESS for each bca axis
the dudi of the bca on which cross-validation should be done
logical to display a progress bar during computations (see the progress package)
further arguments passed to or from other methods
Jean Thioulouse
This function does a cross-validation of the row coordinates of a dudi. Each row is removed from the table one at a time, and its coordinates are computed by projection of this row in the analysis of the table with the removed row. This can be used to check the sensitivity of an analysis to outliers. The cross-validated and original coordinates can be compared with the s.match function (see example).
loocv.between, loocv.discrimin, suprow, s.match
data(meaudret)
envpca <- dudi.pca(meaudret$env, scannf = FALSE, nf = 3)
xvpca <- loocv(envpca)
s.match(envpca$li, xvpca$XValCoord)
Run the code above in your browser using DataLab