Population Pyramids – R for Social Good

library(tidyverse)
library(tidycensus)
library(plotly)

I am working through creating Population Pyramids using the tidycensus package. Also here is another tutorial on Pyramid Plots.

The graph below is copy of the example from the article; the only change is I changed to Polk County Florida, and will add some code comments to help me understand how the code works.

polk <- get_estimates(geography = "county", 
                             product = "characteristics", 
                             breakdown = c("SEX", "AGEGROUP", "HISP"),  
                             breakdown_labels = TRUE, 
                             state = "FL", 
                             county = "Polk")

Review the data structure.

dplyr::glimpse (polk)

Rows: 114
Columns: 7
$ GEOID    <chr> "12105", "12105", "12105", "12105", "12105", "12105", "12105"…
$ NAME     <chr> "Polk County, Florida", "Polk County, Florida", "Polk County,…
$ year     <int> 2022, 2022, 2022, 2022, 2022, 2022, 2022, 2022, 2022, 2022, 2…
$ SEX      <chr> "Male", "Male", "Male", "Male", "Male", "Male", "Male", "Male…
$ AGEGROUP <fct> All ages, All ages, All ages, Age 0 to 4 years, Age 0 to 4 ye…
$ HISP     <chr> "Both Hispanic Origins", "Non-Hispanic", "Hispanic", "Both Hi…
$ value    <dbl> 389036, 273105, 115931, 22146, 13638, 8508, 24387, 15188, 919…

head(polk)

# A tibble: 6 × 7
  GEOID NAME                  year SEX   AGEGROUP         HISP             value
  <chr> <chr>                <int> <chr> <fct>            <chr>            <dbl>
1 12105 Polk County, Florida  2022 Male  All ages         Both Hispanic … 389036
2 12105 Polk County, Florida  2022 Male  All ages         Non-Hispanic    273105
3 12105 Polk County, Florida  2022 Male  All ages         Hispanic        115931
4 12105 Polk County, Florida  2022 Male  Age 0 to 4 years Both Hispanic …  22146
5 12105 Polk County, Florida  2022 Male  Age 0 to 4 years Non-Hispanic     13638
6 12105 Polk County, Florida  2022 Male  Age 0 to 4 years Hispanic          8508

compare <- filter(polk, str_detect(AGEGROUP, "^Age"), #exclude all ages
                  HISP != "Both Hispanic Origins", #this helps us segement by eth rather than totals
                  SEX != "Both sexes") %>% #again we want to segement by eth and sex. 
  mutate(value = ifelse(SEX == "Male", -value, value)) #I think this used for plotting and using the abolute value for the pyramid plot. 

ggplot(compare, aes(x = AGEGROUP, y = value, fill = SEX)) + 
  geom_bar(stat = "identity", width = 1) + 
  theme_minimal(base_family = "Roboto") + 
  scale_y_continuous(labels = function(y) paste0(abs(y / 1000), "k")) +  #I have no idea what abs is. Is is abreivation or aboslute value. 
  scale_x_discrete(labels = function(x) gsub("Age | years", "", x)) + 
  scale_fill_manual(values = c("darkred", "navy")) + 
  coord_flip() + 
  facet_wrap(~HISP) + 
  labs(x = "", 
       y = "2022 Census Bureau population estimate", 
       title = "Population structure by Hispanic origin", 
       subtitle = "Polk County, Florida", 
       fill = "", 
       caption = "Data source: US Census Bureau population estimates & tidycensus R package")

Here is how Chat GPT3 Answered my questions.

Lets riff on this and get a Population Pyramid based on just SEX.

non_eth <- filter(polk, str_detect(AGEGROUP, "^Age"), #exclude all ages
                  HISP == "Both Hispanic Origins", #this helps us segement by eth rather than totals
                  SEX != "Both sexes") %>% #again we want to segement by eth and sex. 
  mutate(value = ifelse(SEX == "Male", -value, value)) #I think this used for plotting and using the abolute value for t

Again, review the data and its structure.

dplyr::glimpse (non_eth)

Rows: 36
Columns: 7
$ GEOID    <chr> "12105", "12105", "12105", "12105", "12105", "12105", "12105"…
$ NAME     <chr> "Polk County, Florida", "Polk County, Florida", "Polk County,…
$ year     <int> 2022, 2022, 2022, 2022, 2022, 2022, 2022, 2022, 2022, 2022, 2…
$ SEX      <chr> "Male", "Male", "Male", "Male", "Male", "Male", "Male", "Male…
$ AGEGROUP <fct> Age 0 to 4 years, Age 5 to 9 years, Age 10 to 14 years, Age 1…
$ HISP     <chr> "Both Hispanic Origins", "Both Hispanic Origins", "Both Hispa…
$ value    <dbl> -22146, -24387, -25485, -25844, -25352, -24998, -27862, -2627…

head(non_eth)

# A tibble: 6 × 7
  GEOID NAME                  year SEX   AGEGROUP           HISP           value
  <chr> <chr>                <int> <chr> <fct>              <chr>          <dbl>
1 12105 Polk County, Florida  2022 Male  Age 0 to 4 years   Both Hispani… -22146
2 12105 Polk County, Florida  2022 Male  Age 5 to 9 years   Both Hispani… -24387
3 12105 Polk County, Florida  2022 Male  Age 10 to 14 years Both Hispani… -25485
4 12105 Polk County, Florida  2022 Male  Age 15 to 19 years Both Hispani… -25844
5 12105 Polk County, Florida  2022 Male  Age 20 to 24 years Both Hispani… -25352
6 12105 Polk County, Florida  2022 Male  Age 25 to 29 years Both Hispani… -24998

polk_sex <- ggplot(non_eth, aes(x = AGEGROUP, y = value, fill = SEX)) + 
  geom_bar(stat = "identity", width = .85) + 
  theme_minimal(base_family = "Roboto") + 
  scale_y_continuous(labels = function(y) paste0(abs(y / 1000), "k")) +  #I have no idea what abs is. Is is abreivation or aboslute value. 
  scale_x_discrete(labels = function(x) gsub("Age | years", "", x)) + 
  scale_fill_manual(values = c("darkred", "navy")) + 
  coord_flip() + 
  labs(x = "", 
       y = "2022 Census Bureau population estimate", 
       title = "Population structure by SEX", 
       subtitle = "Polk County, Florida", 
       fill = "", 
       caption = "Data source: US Census Bureau population estimates & tidycensus R package")

polk_sex

I updated the width of the bars for this plot. I think it looks better and it helps me see the pattern.

I think it may be helpful to make the plot interactive

interactive_plot <- plotly::ggplotly(polk_sex)

interactive_plot

About