Example EDA for team projects

Exploring the penguins dataset

Import data

library(tidyverse)
library(palmerpenguins)

Univariate distributions

Categorical variables

ggplot(data = penguins, mapping = aes(x = species)) +
  geom_bar()

ggplot(data = penguins, mapping = aes(x = island)) +
  geom_bar()

ggplot(data = penguins, mapping = aes(x = sex)) +
  geom_bar()

Species

• Three types
• Adelie most represented, followed by Gentoo and Chinstrap

Island

• Biscoe most frequent, then Dream and Torgersen
• Any connection between species and island?

Sex

• Roughly equal number of males and females
• Some NA values - what to do with those?

Continuous variables

ggplot(data = penguins, mapping = aes(x = bill_length_mm)) +
  geom_histogram()

ggplot(data = penguins, mapping = aes(x = bill_length_mm)) +
  geom_histogram(binwidth = 1)

ggplot(data = penguins, mapping = aes(x = bill_depth_mm)) +
  geom_histogram()

ggplot(data = penguins, mapping = aes(x = bill_depth_mm)) +
  geom_histogram(binwidth = 0.5)

ggplot(data = penguins, mapping = aes(x = flipper_length_mm)) +
  geom_histogram()

ggplot(data = penguins, mapping = aes(x = flipper_length_mm)) +
  geom_histogram(binwidth = 2)

ggplot(data = penguins, mapping = aes(x = body_mass_g)) +
  geom_histogram()

ggplot(data = penguins, mapping = aes(x = body_mass_g)) +
  geom_histogram(binwidth = 100)

• Bill length - bimodal distribution, slightly long right-tail
• Bill depth - bimodal, but higher peak around 18 mm
• Flipper length - bimodal, much higher peak around 190 (unusually large number right at that value), smaller hump around 210
• Body mass - unimodal, right-skewed

Bivariate comparisons

Species and island

Three values each - any geographic trends where species are found?

ggplot(data = penguins, mapping = aes(x = species, fill = island)) +
  geom_bar(position = "fill")

ggplot(data = penguins, mapping = aes(x = island, fill = species)) +
  geom_bar(position = "fill")

• Torgersen only contains Adelie species

• Dream is roughly 50/50 Adelie and Chinstrap

• Biscoe is 75/25 Gentoo and Adelie

• Only Adelie penguins were observed on all three islands

Biological measurements vs. species

Are these species substantively different in terms of size?

ggplot(data = penguins, mapping = aes(x = bill_length_mm)) +
  geom_histogram() +
  facet_wrap(facets = vars(species))

ggplot(data = penguins, mapping = aes(x = bill_length_mm)) +
  geom_histogram() +
  facet_wrap(facets = vars(species), scales = "free_y")

• Adelie penguins have shorter bills than Gentoo or Chinstrap

ggplot(data = penguins, mapping = aes(x = body_mass_g)) +
  geom_histogram() +
  facet_wrap(facets = vars(species))

ggplot(data = penguins, mapping = aes(x = body_mass_g)) +
  geom_histogram() +
  facet_wrap(facets = vars(species), scales = "free_y")

• Gentoo penguins are chonkier than the other two species (higher typical mass)

ggplot(data = penguins, mapping = aes(x = flipper_length_mm)) +
  geom_histogram() +
  facet_wrap(facets = vars(species))

ggplot(data = penguins, mapping = aes(x = flipper_length_mm)) +
  geom_histogram() +
  facet_wrap(facets = vars(species), scales = "free_y")

• Gentoo penguins tend to have longer flippers
• Seems like Gentoos are just bigger overall

Body mass vs. flipper size

ggplot(data = penguins, mapping = aes(x = flipper_length_mm, y = body_mass_g)) +
  geom_point()

ggplot(data = penguins, mapping = aes(x = flipper_length_mm, y = body_mass_g)) +
  geom_point(mapping = aes(color = species))

ggplot(data = penguins, mapping = aes(x = flipper_length_mm, y = body_mass_g)) +
  geom_point(mapping = aes(color = species)) +
  facet_wrap(facets = vars(species))

ggplot(data = penguins, mapping = aes(x = flipper_length_mm, y = body_mass_g)) +
  geom_point(mapping = aes(color = species)) +
  geom_smooth(method = "lm") +
  facet_wrap(facets = vars(species))

• Positive relationship between flipper length and body mass
• Gentoo penguins are larger than the other two species
• Not sure if trends are different across species or just different in scale