Make a Map Function

October 8, 2024

Make a Map Function

What is a Function

• A reusable piece of code that takes some inputs, performs some operations and returns an output
• Elements
  • Inputs (arguments)
  • Operations (code block)
  • Output (return value)
• Like cooking (to produce a dish, you have a recipe and ingredients)
• Don’t repeat yourself (DRY)

Simple Example

# Define the function
square_number <- function(x) {
  result <- x^2
  return(result)
}

# Use the function
square_number(55)

[1] 3025

Tidy Evaluation

In the Tidyverse, ridy evaluation allows you to reference a column without using quotes or referencing the data frame

democracy |> filter(polyarchy > .5)

vs. 

democracy[democracy$polyarchy > .5, ]

or

democracy[democracy[["polyarchy"]] > .5, ]

Tidy Evaluation

• Tidy evaluation lets us refer to column names without quotes, which is great for interactive use.
• But it can be tricky for programming because dplyr and ggplot2 assume that any argument passed to them is already a column name.
• When we pass a variable containing a column name (e.g., a string), R doesn’t automatically evaluate the variable to find the column name.
• We need to explicitly tell R to look inside the variable and treat it as a column.

Tidy Evaluation

• To do this we use {{}} in functions
• For Shiny apps and strings we will use .data[[]]
• Can also use !!sym() in some contexts
• Depricated or less preferred methods include get(), eval(parse()),aes_string() and !!enquo()

Map Function

Map Function

source("functions/wb-maps.R", local = knitr::knit_global())

create_map(var_id = "SL.TLF.CACT.FE.ZS", 
           title= "Female Labor Force Participation", 
           legend_title = "FLFP %", 
           theme = "inferno", 
           direction = -1)

Map Function

library(rnaturalearth)
library(dplyr)
library(ggplot2)
library(ggthemes)
library(wbstats)

create_map <- function(var_id, title, legend_title, theme, direction){

ne_countries(scale = "medium", returnclass = "sf") |> 
  left_join(
    wb_data(var_id, mrnev = 1), 
    join_by(iso_a3 == iso3c)
  ) |> 
  filter(name != "Antarctica") |>  
  ggplot() + 
  geom_sf(aes(fill = .data[[var_id]])) + 
  labs(
    title =  title, 
    fill = legend_title, 
    caption = "Source: World Bank Development Indicators"
    ) +
  theme_map() +
  theme(
    plot.title = element_text(face = "bold"),
  ) +
  scale_fill_viridis_c( 
    option = theme, 
    direction = direction 
    )
}

Map Function

Save the source code in a folder…

source("functions/wb-maps.R", local = knitr::knit_global())

and call in your document…

create_map(var_id = "SL.TLF.CACT.FE.ZS", 
           title= "Female Labor Force Participation", 
           legend_title = "FLFP %", 
           theme = "inferno", 
           direction = -1)

Your Turn

• Take the code from module 3.1, paste in an R script and save in a folder
• Call source() and then the function in your Quarto document
• Change the values for the five parameters
• Go back to the code and add a new parameter (like subtitle)
• Try calling the function again
• Write something about your map and render the Quarto document

10:00

Democracy Function

• Can you make a map function for a democracy indicator?
• Come up with a plan
  • What will the name of the function be?
  • What will the parameters be?
  • What would the
• Take democracy code from last class and implement
• Add more parameters if desired

10:00

Other Functions

• What other visualizations could you “functionize”?
• Line charts?
• Bar charts?
• Scatter plots?
• Try one!

Group and Summarize

# var_id = any World Bank indicator
# mrv = number of most recent values
# group_var = column to group by (country or date)

mean_wb_data <- function(var_id, mrv, group_var){
  wb_data(var_id, mrv = mrv) |> 
    group_by( {{ group_var }}) |> 
    summarize(mean_value = mean(.data[[var_id]], na.rm = TRUE))
}

# Example call
mean_wb_data("SL.TLF.CACT.FE.ZS", 3, date)

Line Chart

create_line_chart <- function(data, x_var, y_var, color_var, title, x_label,
                              y_label, legend_title, color_palette = "turbo",
                              palette_end = 0.8) {
  
  ggplot(data, aes(x = {{ x_var }}, y = {{ y_var }}, color = {{ color_var }})) +
    geom_line() +
    labs(
      title = title,
      x = x_label,
      y = y_label
    ) +
    scale_color_viridis_d(
      name = legend_title,  
      option = color_palette, 
      end = palette_end
    ) +
    theme_minimal()
}

# Example call
create_line_chart(
  data = brics_dem, 
  x_var = year, 
  y_var = v2x_polyarchy, 
  color_var = country_name, 
  title = "Democracy in the BRICS Nations", 
  x_label = "Year", 
  y_label = "Polyarchy Score",
  legend_title = "Index"
)