- orthog: Internal modification
-
New function: xinterp (Resampling of spectra by nterpolation methods)
-
Modification of functions
- scordis and odis: cutoff value calculation has changed
- Arguments of outsdod, pca_rob and pls_rob have changed
-
Function cpplsr has been removed and replaced by aicplsr
-
Internal modifications in functions dfplsr_cov, dfplsr_div
-
Modification of function cpplsr. See help page.
-
Internal modifications in functions dfplsr_cov, dfplsr_div and seldesign
-
Correction of small typo bugs in function cppca
-
Improvements and modifications selangle and selcoll; See the corresponding help pages
-
Code cleaning
- Function selcoef was removed (possibly temporrary)
- Modification of selangle
-
Graphic methods for selecting PCA or PLS model dimension
-
selbroken
-
selhorn
-
selkaiser
-
selkarlis
-
selscree
-
selangle
-
selcoef
-
selcoll
-
-
Graphic methods for selecting regression models
- selsign
- selwik
-
Auxiliaray function
- plotsl
-
cvpcackf to cvpca_ckf
-
cvpcaia to cvpca_ia
-
cvpcatri to cvpca_tri
-
cvpcarw to cvpca_rw
-
dfpcadiv to dfpca_div
-
dfplscov to dfpls_cov
-
dfplsdiv to dfpls_div
-
pcasph to pca_sph
-
pcacr to pca_cr
-
pcarob to pca_rob
-
pca: Row "ncomp=0" removed in output "explvar"
-
Code cleaning
- Code cleaning
-
Gram-Schmidt orthogonalization has been added to function pca_nipals, and modified in pca_nipalsna
-
Code cleaning
-
Major changes of function names: For consistency with S3 rules, all "." in the function names have been replaced by "" or "_"
-
Code cleaning
- cvpca.rw: Row wise PCA Cross-validation
- dfpca.div: Degrees of freedom of PCA models
- cppca: Mallows Cp for PCA models
- summ: outputs changed
- dfplsr.cov to dfplsr.cov
- dfplsr.div to dfplsr.div
- plsrcp to cpplsr
-
Degrees of freedom for PLSR1 models
- plsrdf.cov
- plsrdf.div
-
Mallows Cp for PLSR1 models
- plsrcp
-
plotxna: New argument "grid"
-
pls, plsr, plsdalm, plsda: argument "algo" set to NULL (this has no implication)
- datozone
- fitcv to cvfit
-
Missing data imputation (MDI) in a matrix
-
ximput.ia: IA algorithm
-
plotxna: Plotting missing data in a matrix
-
-
CV for PCA models
-
cvpca.ia: PCA CV with IA algorithm
-
cvpca.tri: PCA CV with ekf-TRI algorithm
-
cvpca.trickf: PCA CV with ckf-TRI algorithm
-
-
pca.nipalsna: Argument "gramschmidt" renamed to "gs", argument "maxit"
increased to 200, plus internal changes -
selwold: Argument "typ" renamed to "type"
- datoctane
- xssr: Sum of squared X-residuals from scores and loadings
-
blocksopls, blocksopca: Input "ncomp" can now have components = 0. Thanks to Evelyne Vigneau (Oniris France)
-
dasdod: Meaning of parameter theta has changed (before was 1 - theta)
-
xfit: Internal changes
-
mse: Internal changes
-
err: Internal changes
- selwold: The function has changed. Only vectors are allowed as input. More generic. New plots.
-
kentson to sampks
-
selncomp.wold to selwold (with slight modification)
-
segmcvkfold to segmkf
-
segmcvmc to segmts
-
plotjit: Jittered plot
-
sampdp: Duplex sampling algorithm
-
Argument ncomp added to functions scordis and odis
-
Argument alpha.f (color opacity for points) added function plotxy
-
Argument weights added to function daglm
-
Argument alpha in functions pca.rob and pls.rob has changed of definition (was 1 - alpha before)
-
Output eig in functions pca.eigen, etc. now contains the full set of eigenvalues (not only the first ncomp values)
-
Output explvarx in functions pca and kpca has been renamed to explvar
-
Function sampclas has changed
-
Re-organisation of some outputs in discrimination functions
-
svmr: SVM regression
-
svmc: SVM discrimination (classification)
-
splitpar: Auxiliary function for tuning models