Linear Regression

Linear regression models the relationship between a numeric outcome and one or more predictor variables, fitting a straight line, or hyperplane with multiple predictors, that best describes the relationship. R's lm() function fits linear models with a simple formula syntax, like lm(sales ~ advertising). Checking the model's assumptions, like linearity and constant variance of the errors, matters for trusting its results.