logistic-regression.r

# first, look at the data
iris
head(iris)
summary(iris)
?iris

# next, visualize the data
library(ggplot2) # if not installed, run this code: install.packages("ggplot2")
ggplot(iris, aes(y = Sepal.Length, x = Petal.Length)) +
  geom_point() +
  geom_smooth(method = "lm")

# next, run a regression
model1 <- lm(Sepal.Length ~ Petal.Length, data = iris)
summary(model1)

#*********** Let's try logistic regression ***************
ggplot(iris, aes(y = Species, x = Petal.Length)) +
  geom_point(alpha = .1) +
  geom_smooth(method = "lm") # oops that didn't work

library(tidyverse) # if you don't have it, install it (install.packages("tidyverse"))
iris2 <- iris %>%
  mutate(versicolor = ifelse(.$Species == "versicolor",1,0)) %>%
  filter(Species != "setosa")

ggplot(iris2, aes(y = versicolor, x = Petal.Length)) +
  geom_point() +
  stat_smooth(method = "glm", method.args=list(family="binomial"), se=FALSE) +
  labs(y = "p(versicolor)", title = "Probability of versicolor species\ngiven petal length")
  
# run our logistic regression
model1 <- glm(versicolor ~ Petal.Length, data = iris2, family = "binomial")
summary(model1)