Exploring the Data with Scatter Plots

I created four dynamic scatter plot tables that show the county by county relationships between COVID-19 mortality and :
- public housing assistance rate
- median age
- median income
- higher education attainment

The script below:

1. Divides the data into quintiles by covid-19 mortality rate per 1,000 population
2. Exports the quintile into a csv file for the shiny table
3. Updates variables names into a descriptive format
4. Creates 4 interactive scatter plots of Michigan counties, with customized colors, themes, and fonts
5. Export plots for as jpegs in the portfolio repository

The final analysis is here

Step 1.1

• Load plotly and tidyverse, ggtheme, and fonts #install package if needed

library(plotly) #make it interactive

## Loading required package: ggplot2

## 
## Attaching package: 'plotly'

## The following object is masked from 'package:ggplot2':
## 
##     last_plot

## The following object is masked from 'package:stats':
## 
##     filter

## The following object is masked from 'package:graphics':
## 
##     layout

library(tidyverse) # create the visualizations

## ── Attaching packages
## ───────────────────────────────────────
## tidyverse 1.3.2 ──

## ✔ tibble  3.1.8     ✔ dplyr   1.0.9
## ✔ tidyr   1.2.0     ✔ stringr 1.4.0
## ✔ readr   2.1.2     ✔ forcats 0.5.1
## ✔ purrr   0.3.4     
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks plotly::filter(), stats::filter()
## ✖ dplyr::lag()    masks stats::lag()

library(ggthemes) # preset plot aesthetics

Step 2.1

• create quintiles for an added layer for the scatter plots

covid_quintile <- read.csv("~/STA 518/BrookemWalters-Portfolio/Stats 518 Final Project/data/covid_census.csv") %>% 
  select(population19E, householdsE, median_ageE, median_incomeE, bach_degree_plus_a25E, unemployment_rate, public_assist_rate, percent_asian, percent_black, percent_native, percent_pacific_islander, percent_white, percent_hispanic, County, Total_Deaths, Deaths_Per_Pop_Thousand) %>% 
  mutate(covid_mortality_quintile = ntile(Deaths_Per_Pop_Thousand,5))

1 = lowest Covid deaths per 1,000 5 = highest Covid deaths per 1,000

Step 2.2

• Create a line of code to generate citations

us_cd_deaths <- read_csv("https://raw.githubusercontent.com/CSSEGISandData/COVID-19/master/csse_covid_19_data/csse_covid_19_time_series/time_series_covid19_deaths_US.csv")

## Rows: 3342 Columns: 941
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## chr   (6): iso2, iso3, Admin2, Province_State, Country_Region, Combined_Key
## dbl (935): UID, code3, FIPS, Lat, Long_, Population, 1/22/20, 1/23/20, 1/24/...
## 
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

source_two_date <- rev(names(us_cd_deaths))[1]
print_sources <-  print(paste ("Sources: 1) American Community Survey (ACS), 2016-2020 5-Year estimates, 2) JHU CSSE COVID-19 Data Repository as of", source_two_date)) 

## [1] "Sources: 1) American Community Survey (ACS), 2016-2020 5-Year estimates, 2) JHU CSSE COVID-19 Data Repository as of 8/7/22"

Step 3

• Convert the quintile numbers into factors, so the variable can have discrete formatting in the plots

covid_quintile$covid_mortality_quintile <-  covid_quintile$covid_mortality_quintile %>%  as.factor 

Step 4

• Convert data into tibble for ease of wrangling

covid_quintile <- tibble(covid_quintile)

Step 5

• Export this data for another project, I’ll update the names next and I don’t want this to interfere with my calculations

write.csv(covid_quintile,"~/STA 518/BrookemWalters-Portfolio/Stats 518 Final Project/Bootstrapping/covid_census_bs.csv")

Step 6

• Update variable with descriptive names, the other set of descriptive names caused an error if used before this point due to the symbols used

covid_quintile_chart <-  covid_quintile %>% 
  rename( 
    Population = "population19E",
    Households = "householdsE",
    `Median Age` = "median_ageE",
    `Median Household Income` = "median_incomeE",
    `% A25+ College Grad+` = "bach_degree_plus_a25E",
    `% of HH on Public Assist.` = "public_assist_rate",
    `Unemployment Rate` = "unemployment_rate",
    `% Asian` = "percent_asian",
    `% Black` = "percent_black",    
    `% Native American` = "percent_native", 
    `% Pacific Islander` = "percent_pacific_islander",
    `% White` = "percent_white",            
    `% Hispanic` = "percent_hispanic",
    `Total Covid Deaths` = "Total_Deaths",
    `Covid Deaths Per Thousand` = "Deaths_Per_Pop_Thousand"
  )

Step 7

• Create a color pallet for quintiles

mortal_palette <-  c("#999999", "#00798c", "#66a182", "#edae49", "#dc0000b2")

Steps 8-12

• create a scatter plot showing the relationship between COVID-19 Deaths per 1000, Total COVID-19 Death, Quintiles,
• update the x axis with the desired variable, and update the title to reflect the changes

assist_sp <- covid_quintile_chart %>% 
  ggplot( aes(x = `% of HH on Public Assist.`, y = `Covid Deaths Per Thousand`, size = `Total Covid Deaths`, color=covid_mortality_quintile, text=County)) +
  geom_point(alpha = 6/10) +
  scale_size_continuous(range = c(2, 15)) +
  labs(title = "COVID-19 Mortality Rate by Percent of Households on Assistance",
       subtitle = "Michigan Counties, deaths per 1000",
       caption = print_sources,
       color = "Quintile", size = NULL) +
  scale_color_manual(values = mortal_palette) +
  theme_fivethirtyeight() +
  theme(axis.title = element_text())

#save the static plot to the "plots" folder
jpeg(filename = "~/STA 518/BrookemWalters-Portfolio/Stats 518 Final Project/Plots/assist_sp.jpeg",
      width = 1000, height = 650)
assist_sp 

#display the interactive version
ia_assist_sp <-  ggplotly(assist_sp )

## Warning: plotly.js does not (yet) support horizontal legend items 
## You can track progress here: 
## https://github.com/plotly/plotly.js/issues/53

ia_assist_sp

#static version
age_sp <- covid_quintile_chart %>% 
  ggplot( aes(x = `Median Age`, y = `Covid Deaths Per Thousand`, size = `Total Covid Deaths`, color=covid_mortality_quintile, text=County)) +
  geom_point(alpha = 6/10) +
  scale_size_continuous(range = c(2, 15)) +
  labs(title = "COVID-19 Mortality Rate by Median Age",
       subtitle = "Michigan Counties, deaths per 1000",
       caption = print_sources,
       color = "Quintile", size = NULL) +
  scale_color_manual(values = mortal_palette) +
  theme_fivethirtyeight() +
  theme(axis.title = element_text())

#save the static plot to the "plots" folder
jpeg(filename = "~/STA 518/BrookemWalters-Portfolio/Stats 518 Final Project/Plots/age_sp.jpeg",
     width = 1000, height = 650)
age_sp 

#display the interactive version
ia_age_sp <-ggplotly(age_sp)

## Warning: plotly.js does not (yet) support horizontal legend items 
## You can track progress here: 
## https://github.com/plotly/plotly.js/issues/53

ia_age_sp

ed_sp <- covid_quintile_chart %>% 
  ggplot( aes(x = `% A25+ College Grad+`, y = `Covid Deaths Per Thousand`, size = `Total Covid Deaths`, color=covid_mortality_quintile, text=County)) +
  geom_point(alpha = 6/10) +
  scale_size_continuous(range = c(2, 15)) +
  labs(title = "COVID-19 Mortality Rate by Higher Education Attainment",
      subtitle = "Michigan Counties, deaths per 1000",
       caption = print_sources,
       color = "Quintile", size = NULL) +
  scale_color_manual(values = mortal_palette) +
  theme_fivethirtyeight() +
  theme(axis.title = element_text())

#save the static plot to the "plots" folder
jpeg(filename = "~/STA 518/BrookemWalters-Portfolio/Stats 518 Final Project/Plots/ed_sp.jpeg",
      width = 1000, height = 650)
ed_sp

#display the interactive version
ia_ed_sp <- ggplotly(ed_sp )

## Warning: plotly.js does not (yet) support horizontal legend items 
## You can track progress here: 
## https://github.com/plotly/plotly.js/issues/53

ia_ed_sp

income_sp <- covid_quintile_chart %>% 
  ggplot( aes(x = `Median Household Income`, y = `Covid Deaths Per Thousand`, size = `Total Covid Deaths`, color= covid_mortality_quintile, text=County)) +
  geom_point(alpha = 6/10) +
  scale_size_continuous(range = c(2, 15)) +
  scale_x_continuous(labels=scales::dollar_format()) +
 labs(title ="COVID-19 Mortality Rate by Median Household Income",
       subtitle = "Michigan Counties, deaths per 1000",
       caption = print_sources,
       color = "Quintile", size = NULL) +
  scale_color_manual(values = mortal_palette) +
  theme_fivethirtyeight() +
  theme(axis.title = element_text())

#save the static plot to the "plots" folder
jpeg(filename = "~/STA 518/BrookemWalters-Portfolio/Stats 518 Final Project/Plots/income_sp.jpeg",
      width = 1000, height = 650)
income_sp

#display the interactive version
ia_income_sp <- ggplotly(income_sp )

## Warning: plotly.js does not (yet) support horizontal legend items 
## You can track progress here: 
## https://github.com/plotly/plotly.js/issues/53

ia_income_sp

Part Four: Making it Shiny

—Back to Table of Contents—