A quick rewrite of Shaun Jackman's original `make` lesson, written for Rich Fitzjohn's `remake`

functions.r

make_plot <- function(mydata){
  qplot(Length, Freq, data=mydata)
}

remake.yml

packages:
  - ggplot2
  - rmarkdown
  
sources:
  - functions.R

targets:
  all: 
    depends:
      - report.html

  words.txt:
    command: file.copy(from = "/usr/share/dict/words", to = target_name)

  words:
    command: readLines("words.txt")
    
  Length:
    command: nchar(words)
  
  hist_dat:
    command: table(Length)
    
  hist_df:
    command: as.data.frame(hist_dat)
  
  histogram.png:
    command: make_plot(hist_df)
    plot: true
    
  report.md:
    knitr: true
    depends:
      - histogram.png
      - hist_df

  report.html:
    command: render("report.md")

report.Rmd

---
title: "English Word lengths"
author: "Jenny Bryan"
date: "`r format(Sys.time(), '%d %B, %Y')`"
output:
  html_document:
    keep_md: yes
---

On most *nix systems, the file `/usr/share/dict/words` contains a bunch of words. On my machine, it contains `r sum(hist_df$Freq)` words.

I computed the length of each word, i.e. the number of characters, and tabulated how many words consist of 1 character, 2 characters, etc.

The most frequent word length is `r with(hist_df, Length[which.max(Freq)])`.

Here is a histogram of word lengths.

![*Fig. 1* A histogram of English word lengths](histogram.png)

Yiou can get a handy diagram with remake::diagram(); that might be the most useful thing for teaching.