predict.mvr {pls} | R Documentation |
Prediction for MVR (PCR, PLSR) models. New responses or scores are predicted using a fitted model and a new matrix of observations.
## S3 method for class 'mvr': predict(object, newdata, comps = 1:object$ncomp, type = c("response", "scores"), cumulative = TRUE, ...)
object |
an mvr object. The fitted model |
newdata |
a data frame. The new data. If missing, the training data is used. |
comps |
vector of positive integers. The components to use in the prediction. See below. |
type |
character. Whether to predict scores or response values |
cumulative |
logical. How the elements of comps are
interpreted. Ignored if type is "scores". See below |
... |
further arguments. Currently not used |
When type
is "response" (default), predicted response
values are returned. If cumulative
is TRUE
, the
elements of comps
are interpreted cumulatively,
i.e. predictions for models with comps[1]
components,
comps[2]
components, etc., are returned. Otherwise, predicted
response values for a single model with the exact components in
comps
are returned.
When type
is "scores", predicted score values are
returned for the components given in comps
.
When type
is "response", a three dimensional array of
predicted response values is returned. The dimensions correspond to
the observations, the response variables and the model sizes,
respectively.
When type
is "scores", a score matrix is returned.
Ron Wehnrens and Bjørn-Helge Mevik
mvr
, summary.mvr
,
coef.mvr
, plot.mvr
data(NIR) nir.mvr <- mvr(y ~ X, ncomp = 5, data = NIR[NIR$train,]) ## Predicted responses for models with 1, 2, 3 and 4 components pred.resp <- predict(nir.mvr, comps = 1:4, newdata = NIR[!NIR$train,]) ## Predicted responses for a single model with components 1, 2, 3, 4 predict(nir.mvr, comps = 1:4, cumulative = FALSE, newdata = NIR[!NIR$train,]) ## Predicted scores predict(nir.mvr, comps = 1:3, type = "scores", newdata = NIR[!NIR$train,])