penguins_analysis.qmd

# Loading packages
```{r}
library(tidyverse)
library(palmerpenguins)
```

```{r}
test_variable <- 5
penguins_df <- penguins
```

```{r}
cleaned_penguins_df <- penguins_df |>
  mutate(age = 2025 - year) 
  # |>
  #filter(flipper_length_mm > 200) |>
  # group_by(species) |>
  # summarize(num_penguins = n(), mean_body_mass = mean(body_mass_g, na.rm = TRUE))
```

```{r}
mean_bill_length <- cleaned_penguins_df |> group_by(species) |> summarize(num_penguins = n(), mean_bill_length = mean(bill_length_mm, na.rm = TRUE))
mean_bill_length
```
```{r}
penguins_df |>
  count(species, island) 

prop_species_df <- penguins_df |> group_by(species, island) |>
  mutate(n_per_species_and_island = n()) |>
  group_by(island) |>
  mutate(n_per_island = n(), prop_species = n_per_species_and_island / n_per_island) 
```
```{r}
ggplot(data=penguins_df,
       mapping = aes(x=flipper_length_mm, y=bill_length_mm)) +
  geom_point(size=4, alpha=0.5) +
  labs(x="Flipper length (mm)", y="Bill Length (mm)", title="Correlation between flipper length and bill length") +
  geom_smooth() +
  theme_classic()
```