Animating the Datasaurus dataset as part of the 30 Day Chart Challenge

Visualization & Pearson’s correlation coefficient

The 30DayChartChallenge is a data visualization community that hosts daily challenges for April. Today’s challenge involves dinosaurs. I used the Datasaurus package to create an animated visualization to demonstrate the importance of graphing data, and the effects of outliers on statistical properties. In these datasets, the mean and the standard deviation of \(x\) and \(y\) remain almost the same, but the distribution of observation pairs goes through obvious drastic changes.

Measures of Central Tendency

library(datasauRus) 
library(knitr)
library(tidyverse)

datasaurus <- datasaurus_dozen
group_datasaurus <- datasaurus %>%
  summarize(avg = mean(x), 
            stdev = sd(x), 
            avg_y = mean(y), 
            stdev_y = sd(y))

knitr::kable(group_datasaurus,
             col.names = c('Mean(x)','SD(x)','Mean(y)','SD(y)'))

Mean(x)	SD(x)	Mean(y)	SD(y)
54.2657	16.713	47.8351	26.84777

Mean & Standard Deviation of x and y remain equal for each dataset

## -------------------------------------- ##
## Day 19: Dinosaurs; 30DayChartChallenge ##
## -------------------------------------- ##
## Loading libs                           ##
library(datasauRus)                       ##
library(tidyverse)                        ##
library(gganimate)                        ##                                                    
library(showtext)                         ##
## -------------------------------------- ##

# Setting custom theme for ggplot2 graphics

clrs <- c( # Creating custom color palette
  "#e5b858", #  yellow
  "#B72025", #  red
  "#7C225C", #  purple
  "#d3d6d9", #  dark gray
  "#f9f9f9", #  light gray
  "#394DAA") #  blue

# Import Fonts (google name, R name)
font_add_google("Source Code Pro","source_code")
font_add_google("Montserrat","mont")
showtext_auto() # enable showtext font rendering

theme_blue <- function() {
  theme_minimal(base_family = "mont") +
  theme(
    plot.title = element_text(family = "mont", face = "bold", size = rel(1.35)),
    plot.subtitle = element_text(family = "mont", colour = "royalblue", face = "plain", size = rel(1)),
    plot.background = element_rect(fill = clrs[5], colour = clrs[5]),
    panel.border = element_blank(),
    panel.background = element_rect(fill = clrs[5]),
    panel.grid.major.x = element_line(colour = clrs[6], linetype = 3, linewidth = .35),
    panel.grid.minor.x = element_blank(),
    panel.grid.major.y =  element_line(colour = clrs[6], linetype = 3, linewidth = .35),
    panel.grid.minor.y = element_blank(),
    # modify text, axis and colour 4) and 5)
    axis.text = element_text(colour = clrs[6], face = "plain", family = "mont"),
    axis.title = element_text(colour = clrs[6], family = "mont"),
    axis.ticks = element_line(colour = clrs[6]),
    # legend at the bottom 6)
    legend.position = "none"
  )
}

theme_set(theme_blue())

ggplot2::update_geom_defaults("label", list(family = "mont", fontface = "bold"))
ggplot2::update_geom_defaults("text", list(family = "mont", fontface = "plain"))


# Checking min/max for axis transform
datasaurus <- datasaurus_dozen
summary(datasaurus)

dino <- ggplot(datasaurus, aes(x = x, y = y))+
  geom_point(size = 3, alpha = .65, 
             colour = clrs[3]) +
  labs(x = NULL,y = NULL,
       title = "Day 19 of 30 Day Chart Challenge",
       subtitle = "The T-rex",
       caption = "Always check your data graphically") +
  theme_blue() +
  transition_states(states = dataset,
                    transition_length = 2,
                    state_length = 1,
                    wrap = TRUE) +
  ease_aes("linear")

animate(dino, duration = 15, fps = 24, height = 680, width = 720,
        renderer = gifski_renderer())

anim_save(filename = "dino.gif",
          path = here::here("blog","2024","04","dinosaurs","img"))

The Animated Data