In olobiolo/Rdlazer: Resources for Teaching R

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

What is a formula

There are three classes of language elements in R: names, expressions and calls. A formula is a kind of call that describes a relationship between some variables in a schematic way. The formula interface is used more or less consistently across many functions in R, which is convenient once you get used to it. For example, scatter plot and linear regression have he exact same syntax with a formula

An object of class formula consists of terms (the variables) and a tilde operator ~. It is created with bare term names. Character strings can be converted into formulas with formula or as.formula.

y ~ x

formula("y ~ x")

as.formula(paste('y', 'x', sep = ' ~ '))

Interpretation

The parts of a formula on either side of the tilde are commonly referred to as LHS (left-hand side) and RHS (right-hand side). Typically the LHS represents the response whereas the RHS contains predictors. In some cases, one of the sides can be missing to represent "no response" or consist of a period to represent "all other variables".

The formula above would be interpreted as "y depends on x (in some way)". In a model r ~ x + y + z would mean that the response r is shaped by the x, y and z predictors. The + need not actually mean addition as the variables need not be numeric (they can be factors). It is a way of listing multiple terms which can be independent or interact in some way. Formulas can contain other operators to represent these interactions, depending on the context; this is usually well documented.

Evaluation

Formulas are used within functions and it is the calling function that specifies the environment where the formula will be evaluated, i.e. where the terms will be scoped. The environment can also be a list or a data frame (see the vignette on non-standard evaluation for more details).

WARNING: Formulas are greedy and if they fail to find their terms in the specified frame, they will scope them further up enclosing frames.

olobiolo/Rdlazer documentation built on Aug. 6, 2022, 11:37 a.m.