Every Tom, Dick, and Harry

Writing Functions in R

CVC 2016

require(mosaic)
require(babynames)

What ever happened to Jane (and Mary)?

babynames %>% 
  filter(name %in% c("Mary", "Jane"), sex == "F") %>%
  ggplot() +
  geom_line(aes(x = year, y = prop, color = name))

What about Tom, Dick, and Harry?

We would like to generalize our code to handle any set of names we want to compare. In R we do this by writing functions.

We begin with our two questions:

  1. What do we want R to do?

    • This determines the name of the function (and what it does/returns)
    • NameTrendPlot() might be a good name for the function

  2. What does it need to know to do this?

    • This determines the arguemnt list (list of inputs) to the function
    • The function will need to know which names to track.
    • We should probably have an argument to restrict to just one sex.

Creating a function in R

The general form of a fucntion definition is

functionName <- function( arg1 = default1, arg2 = default2, ... ) {
  # do stuff here
  # result of last line is returned.  
  # Can also use explicit return().
}

What about Tom, Dick, and Harry?

NameTrend <- function(names, gender) {
  babynames %>% 
  filter(name %in% names, sex == gender) %>%
  ggplot() +
  geom_line(aes(x = year, y = prop, color = name))
}

What about Tom, Dick, and Harry?

NameTrend(c("Tom", "Dick", "Joe"), gender = "M")

Adding more arguments

We can arguments to control more features of the plot.

NameTrend2 <- function(names, gender = c("M", "F"), linewidth = 1) {
  babynames %>%
    filter(name %in% names, sex %in% gender) %>%
    ggplot() +
    geom_line(aes(x = year, y = prop, color = name, linetype = sex),
              size = linewidth)
}
NameTrend2( c("Tom", "Dick", "Harry"), gender = "M", linewidth = 3)

NameTrend2( c("Hillary", "Beverly"), linewidth = 1)