jennybc

library(testthat)
library(reshape2) # installed from github
                  # SHA 096afb80192f163ced5d001cc416acc1d9a85d12

## working with example from test-melt.r line 132 -->

df <- data.frame(
  id=1:2,
  f1=factor(c("a", "b")),
  f2=factor(c("b", "a"))

jennybc

library(plyr)
library(ggplot2)

gDat <- read.delim("gapminderDataFiveYear.txt")

d_ply(gDat, ~ country, function(z) {
  the_country <- z$country[1]
  the_continent <- z$continent[1]
  
  the_filename <- paste0(the_continent, "_", the_country, ".tsv")

jennybc

library(plyr)

## what's the best way to get tabulation of (days31, oystersOK) into a
## data.frame AND get a grand total row reflecting the expected total of 12
## months?

## related discussion has occured before here:
## http://stackoverflow.com/questions/4946873/how-do-i-add-a-row-to-a-data-frame-with-totals

## related issue open on dplyr:

jennybc

library(dplyr)

## accept default behavior, i.e. stringsAsFactors = TRUE
## downside is that 'name' really shouldn't be a factor
df <- data.frame(name = c("jenny", "jim", "reed", "lorrie"),
                 gender = c("female", "male", "male", "female"))
df %>% group_by(gender) %>% summarize(count = n()) # this works

## protect a single character variable with I()
df <- data.frame(name = I(c("jenny", "jim", "reed", "lorrie")),

jennybc

---
title: "Wrapper function to open/close graphics device"
author: "Jenny Bryan"
date: "25 July, 2014"
output:
  html_document:
    keep_md: TRUE
---
  
```{r setup, include = FALSE, cache = FALSE}

jennybc

```{r setup, include = FALSE, cache = FALSE}
knitr::opts_chunk$set(error = TRUE)
```

jennybc

Inspired by recent @johnmyleswhite tweets on indexing (though they focus more on NA stuff)

My standard advice to R novices is "don't worry about numeric vs integer too much, for now", but I just created an indexing example that scares me.

I did not realize it was sooo easy to create this gotcha when indexing with numeric (which I know is not a great idea).

jennybc

Demo errors in knitted doc vs. other

Jenny Bryan
29 September, 2014

Context: In a recent lesson, I claim (with some justification) that raising an error inside a function via if() ... stop() often gives more info than using stopifnot(). This is partly because you get to write your own error message. But also because you learn the name of the function that threw the error. Or so I thought.

Here's a toy function to generate an error.

jennybc

suppressPackageStartupMessages(library(dplyr))

Get word counts for Character * Chapter in LOTR trilogy

if(!file.exists("lotr_clean.tsv")) {
  download.file(paste0("https://raw.githubusercontent.com/jennybc/",
 "lotr/master/lotr_clean.tsv"), 

jennybc

#' Digest the Link header in a paginated result.
#' 
#' Converts the Link header from a monolithic string to a usable data.frame.
#' 
#' The GitHub API automatically paginates when the number of requested items 
#' exceeds the number of items per page. When this occurs, the result returned 
#' by the server will include a Link header that provides the URLs for other 
#' pages of results, such as the next page and the last page. These assorted 
#' URLs are catenated in a single string and this function converts that 
#' information into a data.frame that is useful for traversing the pages.