Prediction on the test set

This function takes the test set instances and the model trained by plm.trainer in order to predict the classes.

plm.predictor(feat, label, data.split = NULL, models.list, model.mat,
  model.type)

Arguments

feat

features object
label

label object
data.split

filename containing the test samples or list of test instances produced by data.splitter(), defaults to NULL leading to testing on the complete dataset
model.mat

model matrix needed to rebuild the model
model.type

string, type of the model that was trained
model

model object trained by plm.trainer

Value

list containing the precitions

  • $pred;

  • $mat