This class represents the structure we see as a relevant variable. A relevant variable in our case is a variable for which we want to estimate its conditional distribution. This class helps organizing / creating its formula (the parametric form of the covariates that could predict this relevant variable).

RelevantVariable

Format

An object of class R6ClassGenerator of length 24.

Methods

initialize(formula, family)

Creates a new RelevantVariable object. It creates this object according to a formula (the y ~ x1 + x2 representation we expect to predict this relevant variable) and the family (gaussian if it is a double / integer value, and binomial if it is a binary variable) @param formula formula the formula to use for predicting this relevant variable. @param family string a string representing the type of relevant variable this is.

get_formula_string(X = NULL, Y = NULL)

Returns a string representation as a formula of the current relevant variable and its predictors. @param X string (default = NULL) the variables to use as the predicting variables in this formula. If NULL, it will use the X of the current relevant variable. @param Y string (default = NULL) the variables to use as the outcome variable in this formula. If NULL, it will use the X of the current relevant variable.

parse_formula(formula)

Converts a formula to its separate X and Y components. @param formula formula the formula to convert. @return list a list with two attributes: the Y which is the output of the formula and the X the independent variables in this formula.

getFamilty

Active method. Returns the family provided on initialization. @return string the family.

getX

Active method. Returns the X component of the formula provided on initialization. @return list the X component of the formula

getY

Active method. Returns the Y component of the formula provided on initialization. @return string the outcome variable name

getValidity

Active method. Checks the validity of the RelevantVariable instance. This method is called automatically on initialization. @return boolean true if everything is valid, or it throws if not valid.