README.md

HSDS

The goal of HSDS is to make all the data sets of the book “A Handbook of Small Data Sets” (1994) of David J. Hand available. These data sets are particularly useful to demonstrate examples of function or statistical tests, but also to teach about statistics and R.

All data sets are already available individually at this repo: https://github.com/JedStephens/Handbook-of-Small-Data-Sets/tree/master. However, they are not immediately usable in R, and undocumented. This package aims to solve this issue, and provide clean and documented data sets.

Do you like this package and want to support me ?

Installation

You can install the development version of HSDS like so:

devtools::install_github("ABohynDOE/HSDS")

Available data sets

The book contains more than 500 data sets. For the moment, only some are available. They are summarized in the table below, along with their names, what they contain, their structure, and the type of variables present.

name	Title	Structure	Variables
germin	Germinating seeds	48 × 3	factor(2), numeric(1)
lengths	Guessing lengths	113 × 3	character(1), numeric(2)
darwin	Darwin’s cross-fertilized and self-fertilized plants	30 × 3	factor(1), integer(1), numeric(1)
interval	Intervals between cars on the M1 motorway	41 × 2	character(2)
tearing	Tearing factor for paper	20 × 2	numeric(2)
abrasion	Abrasion loss	30 × 3	numeric(3)
water	Mortality and water hardness	61 × 5	factor(1), numeric(4)
cement	Tensile strength of cement	21 × 2	numeric(2)
weight	Weight gain in rats	40 × 3	factor(2), numeric(1)
chickens	Weight of chickens	24 × 3	factor(2), numeric(1)
flicker	Flicker frequency	27 × 4	factor(3), numeric(1)
chloride	Effect of ammonium chloride on yield	32 × 5	factor(4), numeric(1)

Example

This is a basic example which shows you how to use a data set to make a nice plot:

library(HSDS)
library(ggplot2)

ggplot(germin, aes(x = water, y = seeds, color = box)) +
  geom_boxplot(na.rm = T) +
  theme_bw()

Contributing

We are far from the 500 data sets, so any help is welcome ! If you want to contribute, all raw data sets are already present in the repo (at data-raw/data-files), so feel free to clean one or more… ! If you do so, please respect the following guidelines:

• data sets should be named after the data structure index of the book (available here)

• all variables in the data set should be labelled (using the labelled package for example)

• data sets should be documented using the text from the book