In JustinaTse/bis557: Package Encapsulating work for BIS557

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

library(bis557)
library(palmerpenguins)
library(DT)
data(penguins)
datatable(penguins)

Above is the penguins data set form the {palmerpenguins} package.

Selecting variables and creating a table

library(dplyr)
library(ggplot2)
# The "base" way.
table(penguins$species, penguins$island)
# The "tidy" way.
penguins %>%
  select(species, island) %>%
  table()

Filtering rows

base

table(penguins[penguins$species == "Adelie", "sex"])

tidy

penguins %>%
  filter(species == "Adelie") %>%
  select(sex) %>%
  table()

Building a model

Let's fit the model:

$$ \text{bill_length_mm} \sim \text{species} \, \beta_1 + \text{island} \, \beta_2 + \text{sex} \, \beta_3 + \beta_0 $$

To fit the model we'll run the following code:

fit <- lm(bill_length_mm ~ species + island + sex, data = penguins)
summary(fit)

Plot the residuals

library(ggplot2)
# base
qs <- qqnorm(fit$residuals, plot.it = FALSE)
qsd <- as.data.frame(qs)
ggplot(qsd, aes(x = x, y = y)) + 
  geom_point() +
  ylab("Sample Quantiles") +
  xlab("Theoretical Quantiles") +
  theme_minimal()

# tidy
qqnorm(fit$residuals, plot.it = FALSE) %>%
  as_tibble() %>%
  ggplot(aes(x = x, y = y)) + 
    geom_point() +
    ylab("Sample Quantiles") +
    xlab("Theoretical Quantiles") +
    theme_minimal()