[M]-Linear-Regression.md

Linear Regression

Implicit Assumptions
◻️ Dependent variable is continuous.

Actual Code

1. Run the linear regression model.

Regression-based Forecasting

Linear Regression with Principal Components

Simple Linear Regression

fit <- lm(mpg ~ hp, mtcars)
summary(fit)

Multiple Linear Regression (w/o Interaction)

fit <- lm(Petal.Length ~ Species + Sepal.Length, iris)
summary(fit)

Multiple Linear Regression (w Interaction)

fit <- lm(Petal.Length ~ Species * Sepal.Length + Sepal.Width, iris)
#fit <- lm(Petal.Length ~ Species * Sepal.Length, iris)
summary(fit)

2. Examine its residual plots.

plot(fit, 1)
plot(fit, 2)
abline(a = 0, b = 0)

3. If required, obtain prediction and/or residuals.

iris$predicted <- predict(fit)
iris$residuals <- residuals(fit)