This is the function that carries out the prediction of the new observations.

1 2 |

`object` |
This must be a |

`newdata` |
A list of new observations. The format of this set of data must be the same as the training data, including the order of the variables. |

`...` |
Other arguments to input. |

A matrix of predictions. Since the input `flars`

object may have more than one estimated coefficients, the number of predictions may be more than one set. Each column of the outcome is corresponding to one set of coefficients.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 | ```
library(flars)
library(fda)
## Generate some data.
dataL=data_generation(seed = 1,uncorr = TRUE,nVar = 8,nsamples = 120,
var_type = 'm',cor_type = 3)
## Split the training data and the testing data
nTrain=80
nsamples=120
TrainIdx=seq(nTrain)
TestIdx=seq(nsamples)[-TrainIdx]
fsmTrain=lapply(dataL$x,function(fsmI) fsmI[TrainIdx,,drop=FALSE])
fsmTest=lapply(dataL$x,function(fsmI) fsmI[TestIdx,,drop=FALSE])
yTrain=dataL$y[TrainIdx]
yTest=dataL$y[TestIdx]
## Do the variable selection
out=flars(fsmTrain,yTrain,method='basis',max_selection=9,
normalize='norm',lasso=FALSE)
## Do the prediction
pred=predict(out,newdata = fsmTest)
# apply(pred,2,flars:::rmse,yTest)
``` |

Questions? Problems? Suggestions? Tweet to @rdrrHQ or email at ian@mutexlabs.com.

All documentation is copyright its authors; we didn't write any of that.