rtemis (version 0.79)

rtModSet: rtemis model

Description

Creates an rtemis object of your choice of class. All available class systems are supported, but at this point R6 is the default and most widely tested.

Usage

rtModSet(rtclass = "rtMod", mod = list(), mod.name = character(),
  type = character(), parameters = list(), call = "",
  y.train = numeric(), y.test = numeric(), x.name = character(),
  y.name = character(), xnames = character(),
  bag.resample.rtset = list(), fitted.bag = numeric(),
  fitted = numeric(), fitted.prob = numeric(),
  se.fit.bag = numeric(), se.fit = numeric(), error.train = list(),
  predicted.bag = numeric(), predicted = numeric(),
  predicted.prob = numeric(), se.prediction.bag = numeric(),
  se.prediction = numeric(), error.test = list(), varimp = numeric(),
  question = character(), extra = list())

Arguments

rtclass

String: Which class system to use. Previously "S3", "S4", "RC", "R6". Currently, only "R6", aka "rtMod" is supported.

mod

Trained model

mod.name

Model (algorithm) name

type

"Regression" or "Classification"

parameters

List of model hyperparameters

call

Originating call

y.train

Training y data

y.test

Validation y data

x.name

Name of x data

y.name

Name of y data

xnames

Column names of x

fitted

Fitted values

fitted.prob

Fitted probabilities (Classification only)

se.fit

Standard error of the fit

error.train

Training error

predicted

Predicted values

predicted.prob

Predicted probabilities (Classification only)

se.prediction

Standard error of the prediction

error.test

Validation error

varimp

Variable importance

question

Question the model is attempting to answer

extra

List with algorithm-specific extra fields