stacked PLS regression
model=spls(x,y,lv,obs_split,var_split,(centred))
calibration dataset; a matrix (n x q) and a vector (n x 1) or Div structures
number of latent variables
parameter for the cross-validation:
- an integer: number of random blocs
- two integers [a b]: a random blocs ; b repetitions of the cross-validation
- a vector of dimension n attributing each observation to a bloc (numbers 1,2,...k for k blocs): blocs given by the vector
- a matrix (n x b ) of column-vectors of dimension n attributing each observation to a bloc (numbers 1,2,...k for k blocs): blocs given by each vector, b repetitions of the cross-validation
- 'vnbxx': venitian blinds, xx blocs; ex: 'vnb10' for 10 blocs
- 'blkxx' or 'jckxx':contiguous blocs = Jack knife, xx blocs; ex: 'jck8' for 8 blocs
parameter to classify the variables into groups:
- an integer: number of random blocs
- a vector of dimension q attributing each variable to a bloc (numbers 1,2,...k for m blocs): blocs given by the vector
- 'vnbxx': venitian blinds, xx blocs; ex: 'vnb10' for 10 blocs
- 'blkxx' or 'jckxx':contiguous blocs = Jack knife, xx blocs; ex: 'jck8' for 8 blocs
centering=1 (by default); no centering = 0
the standard errors of calibration aand cross-validation
model.err.d is a matrix (lv x 2); the columns are the rmsec and the rmsecv respectively
y predicted after cross validation for each of the latent variables
model.ypredcv.d is a matrix of dimensions (n x lv)
the b coefficients or regression coefficients
model.b.d is a matrix of dimensions (q x lv)
means of x and y; a vector (q x 1) and a scalar
1 = centred, 0 = not centred