Choosing a graphic form, Fundamental chart types (part I)

Artwork by Allison Horst

Pre-class Prep

Before coming to class, you’ll need to install some packages, download data, and request a US Census Bureau API key. For step-by-step instructions, unfold the following note (collapsed to save space):

Step-by-step pre-class prep instructions

Install required packages

install.packages("chron") # for working with dates / times
install.packages("naniar") # tools for exploring & handing missing data
install.packages("tidytuesdayR") # used to download TidyTuesday data
install.packages("tidycensus") # an R package that allows users to interface with a select number of the US Census Bureau’s data APIs and return tidyverse-ready data frames
install.packages("ggridges") # {ggplot2} extension for creating ridgeline plots
install.packages("gghighlight") # {ggplot2} extension for highlighting geoms
install.packages("ggbeeswarm") # {ggplot2} extension for creating categorical scatter (violin point) plots
install.packages("see") # {ggplot2} extension for model visualization (we'll be using it for it's geom, geom_violindot())
install.packages("scales") # provides the internal scaling infrastructure used by ggplot2, and gives you tools to override the default breaks, labels, transformations and palettes (installed automatically with {ggplot2} or {tidyverse})

Download data

• County-level Lyme disease data from 2001-2020, made available by the CDC (read more on this CDC web page)
  • save the downloaded data (LD-Case-Counts-by-County-01-20.csv) to your class repo at the file path: EDS-240-class-examples/week3/data/LD-Case-Counts-by-County-01-20.csv

Request a US Census Bureau API key

1. To use the {tidycensus} package to access US Census Bureau data, you’ll first need to request and activate an API key. Do so by following these steps:

   1. Request an API key at http://api.census.gov/data/key_signup.html
   2. After a short period of time, you should receive an email with your key (it may be helpful to star/bookmark that email), and click the included link to activate your key (Note: if you get an error when activating your key, close your browser window and try clicking the activation link again)

2. For security purposes, you should never push any API keys to GitHub. To prevent us from accidentally pushing our key(s), we can save it to a separate file which we’ll add to our .gitignore. Do so by following these steps:

1. Create a week3/ folder in the root directory of your class repo

2. Add a file to your week3/ folder called KEYS.R, and save your API key to an object name like this:

~/week3/KEYS.R

censusKEY <- "XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX"

2. Add KEYS.R to your .gitignore

.gitignore

/week3/KEYS.R

3. Save .gitignore (once saved, you should see the newly added KEYS.R file disappear from your Git pane in RStudio, and a modified .gitignore appear). Stage/add, commit, and push this .gitignore file to GitHub

In class, we’ll source our key (by running source(here::here("week3", "KEYS.R"))) into any scripts / Qmd files, where required.

Lecture Materials

Week 3 instruction is broken down into three lessons:

Choosing graphic forms

lecture 3.1 slides

Visualizing distributions

lecture 3.2 slides

Visualizing evolution

lecture 3.3 slides

Today’s lectures are inspired by and draw largely on materials created by a number of other educators, including Albert Cairo and his book The Truthful Art, Meghan Hall and her course, CMU 36-315: Statistical Graphics & Visualization, as well as information curated by Yan Holtz & Conor Healy as part of their From Data to Viz project.

Discussion Materials

Pre-discussion Prep

Before coming to section, please be sure to:

1. Install the following packages:

install.packages("geofacet") # facet data for different geographical regions using panels arranged in shapes that mimic geographic topology

2. (Optional) Install the ColorPick Eyedropper Google Chrome extension, which makes it easy to select color values from webpages.

Background

The U.S. Drought Monitor (USDM) is a collection of measures that allows experts to assess droughts in the United States. It is produced through a partnership between the National Drought Mitigation Center at the University of Nebraska-Lincoln, the United States Department of Agriculture and the National Oceanic and Atmospheric Administration. You can download and explore comprehensive statistics through their data portal.

The TidyTuesday community wrangled and visualized USDM data back in 2021 (2021-07-20, week 30), so we’ll use the {tidytuesdayR} package to import some minimally-tidied data (TidyTuesday always provides the cleaning script that the organizers used to pre-process any of the data provided – these data comprise three separate data sets, which were converted from wide to long format, joined together, and had some columns renamed).

Over the course of the next two discussion sections (week 3 & week 4), you’ll be recreating the following data visualization produced by U.S. Drought Monitor, which you can find on the Droughts in California Wikipedia page:

This week, we’ll focus on the data layer, geometric layer, and scales. Next week, we’ll work on tweaking the theme (all non-data elements) to get it to look just like the U.S. Drought Monitor’s version.

Solution

You’ll get the most out of discussion section if you physically type out the code yourself (rather than copying / pasting)!

Complete code for week 3 discussion

##~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
##                                    setup                                 ----
##~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

#..........................load packages.........................
library(tidyverse)

#..........................import data...........................
tuesdata <- tidytuesdayR::tt_load('2021-07-20')
drought <- tuesdata$drought

##~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
##                            wrangle drought data                          ----
##~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

drought_clean <- drought |>

  # select cols of interest & update names as needed ----
  select(date = valid_start, state_abb, drought_lvl, area_pct) |> 

  # add year, month & day cols using {lubridate} fxns ----
  # NOTE: this step isn't necessary for our plot, but I'm including as examples of how to extract different date elements from a object of class Date using {lubridate}~ ----
  mutate(year = year(date),
         month = month(date, label = TRUE, abbr = TRUE),
         day = day(date)) |>

  # add drought level conditions names ----
  mutate(drought_lvl_long = factor(drought_lvl,
                            levels = c("D4", "D3", "D2", "D1","D0", "None"),
                            labels = c("(D4) Exceptional", "(D3) Extreme",
                                       "(D2) Severe", "(D1) Moderate", "(D0) Abnormally Dry", 
                                       "No Drought"))) |>
  
  # reorder cols ----
  select(date, year, month, day, state_abb, drought_lvl, drought_lvl_long, area_pct)

##~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
##       create stacked area plot of CA drought conditions through time     ----
##~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

drought_clean |> 
  
  # remove drought_lvl "None" & filter for just CA ----
  filter(drought_lvl != "None",
         state_abb == "CA") |> 
  
  # create ggplot ----
  ggplot(mapping = aes(x = date, y = area_pct, fill = drought_lvl_long)) +
  
  # reverse order of groups so level D4 is closest to x-axis ----
  geom_area(position = position_stack(reverse = TRUE)) +
  
  # update colors to match US Drought Monitor (colors selected using ColorPick Eyedropper from original USDM data viz) ----
  scale_fill_manual(values = c("#853904", "#FF0000", "#FFC100", "#FFD965", "#FFFF00")) +
  
  # set x-axis breaks & remove padding between data and x-axis ----
  scale_x_date(breaks = scales::breaks_pretty(n = 10),
               expand = c(0, 0)) +

  # set y-axis breaks & convert values to percentages & & remove padding between data and y-axis----
  scale_y_continuous(breaks = seq(0, 100, by = 10),
                     labels = scales::label_percent(scale = 1),
                     expand = c(0, 0)) +
  
  # add title ----
  labs(title = "Drought area in California") 

##~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
##            BONUS: using {geofacet} to plot data for all states           ----
##~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

# NOTE 2024-01-24: The following code works for Sam, but apparently no one else! I'll keep digging into this and update folks when I have a solution

#......remove HI & AK (no data) and also DC from preset grid.....
mygrid <- geofacet::us_state_grid1 |>
  filter(!code %in% c("DC", "HI", "AK"))

#..............................plot..............................

# NOTE: this takes a minute to render!

drought_clean |> 
  filter(drought_lvl != "None") |> 
  ggplot(aes(x = date, y = area_pct, fill = drought_lvl_long)) +
  geom_area(position = position_stack(reverse = TRUE)) +
  geofacet::facet_geo(~state_abb, grid = mygrid) +
  scale_fill_manual(values = c("#853904", "#FF0000", "#FFC100", "#FFD965", "#FFFF00"))

Assignment Reminders

Assignment Type	Assignment Title	Date Assigned	Date Due
EOC	EOC (week 3)	Mon 01/22/2024	Mon 01/22/2024, 11:55pm PT
HW	Homework Assignment #2	Mon 01/22/2024	Sat 02/03/2024, 11:59pm PT