For each point we have the predicted class / regression value in column response. If the learner predicts probabilities, a column ".prob.response" is added that contains the probability of the predicted class

predict_grid(learners, task, grid_points, expand_range)

Arguments

learners

list of trained learners, each learner belongs to one resampling iteration

task

the task all learners are trained on

grid_points

(int): see sequenize

expand_range

see sequenize