Datasets to accompany the book Learning R

Description

learningr contains datasets that are used in examples in the book “Learning R”.

Author(s)

Richard Cotton richierocks@gmail.com

References

Cotton, R (2013) Learning R O Reilly. ISBN 978-1-4493-5710-8.

Alpe d'Huez

Description

Fastest times for the Alpe d'Huez stage of the Tour de France cycle race, plus some contextual information on year and drug use allegations.

Format

A data frame with the following columns.

Time

Character time of ride in the form M' S".

NumericTime

Numeric time of ride in minutes.

Name

Name of rider.

Year

Year of race.

Nationality

Nationality of rider.

DrugUse

Have allegations of drug use been made against the rider? In alpe_d_huez the values are "Y" or "N"; in alpe_d_huez2 this is a logical vector.

Note

The data was kindly compiled by William Hogan.

The dataset is not guaranteed to be error free. Please double check the data if you want to use it for something serious.

References

William Hogan william.hogan@amd.com compiled the data from http://en.wikipedia.org/wiki/Alpe_d%27Huez. Richard Cotton richierocks@gmail.com made some modifications while importing it into R.

A buggy version of plyr's count function

Description

An old version of the plyr package's count function that fails when you pass it a factor input.

Usage

  buggy_count(df, vars = NULL, wt_var = NULL)

Arguments

Value

A data frame with label and freq columns.

Note

In case the “buggy” part of the name didn't give it away, this is not suitable for real world usage!

See Also

count and table

Examples

## Not run: 
buggy_count(factor()) #oops!

## End(Not run)

Crab tag

Description

Depth and temperature data from a sensor tag attached to an edible crab (Cancer Pangurus) in the North Sea in 2008 and 2009.

Format

A list with 5 elements, as follows. id_block is a list with 2 elements

Firmware Version No

Version of the firmware used in the crab tag.

Firmware Build Level

Build level of the firmware used in the crab tag.

tag_notebook is a list with 5 elements.

Mission Day

Number of days of data.

Last Deployment Date

Date and time that the tag was released into the sea.

Deployed by Host Version

UNKNOWN

Downloaded by Host Version

UNKNOWN

Last Clock Set Date

UNKNOWN

lifetime_notebook is a list with 3 elements.

Tag ID

The unique ID of the tag.

Pressure Range

UNKNOWN

No of sensors

The number of sensors on the tag.

deployment_notebook is a data frame with X columns.

Start

Start date and time of recording.

Stop

Stop date and time of recording.

Logging Rate

UNKNOWN

Pointer

UNKNOWN

PA inc

UNKNOWN

sensors

UNKNOWN

Flags

UNKNOWN

Resolution

UNKNOWN

Fast Rate

UNKNOWN

V1

UNKNOWN

V2

UNKNOWN

daylog is a data frame with X columns.

Mission.Day

Integer number of days since the start of recording.

Date

Date of record.

Max.Temp

Maximum temperature (Celcius) recorded that day.

Min.Temp

Minimum temperature (Celcius) recorded that day.

Max.Depth

Maximum depth (m) recorded that day.

Min.Depth

Minimum depth (m) recorded that day.

Batt.Volts

Voltage of tag battery.

Note

This data was kindly supplied by Ewan Hunter of CEFAS. It is part of a larger dataset consisting of many crabs.

References

Ewan Hunter ewan.hunter@cefas.gov.uk ran the project where the data was collected. A full analysis is in Hunter, E, Eaton, D, Stewart, C, Lawler, A & Smith, M. 2013. Edible crabs "go west": migrations and incubation cycle revealed by electronic tags. https://www.ncbi.nlm.nih.gov/pubmed/23734180. Richard Cotton richierocks@gmail.com made some modifications while importing it into R.

Deer Endocranial Volume

Description

The dataset contains the endocranial volume of 33 red deer (Cervus elaphus), using four different methods of measurement: cardiac tomography, filling the skull with glass beads (yes, the skulls are from dead deer), simply measuring the length, width and height and multiplying the numbers, and using Finarelli's equation. "Endocranial volume" is a proxy for brain size.

Format

A data frame with the following columns.

SkullID

A unique identifier for each red deer.

VolCT

Endocranial volume calculated by cardiac tomography.

VolBead

Endocranial volume calculated by glass beads.

VolLWH

Endocranial volume calculated by length*width*height.

VolFinarelli

Endocranial volume calculated by Finarelli's equation.

VolCT2

A second measurement via cardiac tomography.

VolBead2

A second measurement via glass beads.

VolLWH2

A second measurement via l*w*h.

Note

The data was kindly provided by Corina Logan. Second measurements are provided for several of the deer. Finarelli's equation is used for estimating the brain volume of non-bovid ruminant Artiodactylid species (say that 10 times fast).

\ln(volume(skull)) = 2.6616 * \ln(width(skull)) - 6.2722

References

The dataset was collected by Corina Logan itsme@corinalogan.com. It is stored in the Dryad Digital Repository, doi: 10.5061/dryad.4t7h2. A more full analysis is given in the paper Logan CJ, Clutton-Brock TH. 2013. Validating methods for estimating endocranial volume in individual red deer (Cervus elaphus). Behavioural Processes 92:143-146. doi:10.1016/j.beproc.2012.10.015. http://www.sciencedirect.com/science/article/pii/S037663571200232X Richard Cotton richierocks@gmail.com made some modifications while importing it into R.

English Monarchs

Description

Names, dates and houses of English kings and queens from post-Roman rule (the fifth century) until England invaded Ireland in the Early 13th century.

Format

A data frame with the following columns.

name

Name of monarch(s).

house

Royal house of monarch(s).

start.of.reign

Year they rose to power.

end.of.reign

Year they left power.

domain

Region of England ruled over.

Note

This dataset is a bit messy and ambiguous in places, because history is like that. In fact, the messy parts of the dataset are in general a good indicator that soemthing interesting was happening at the time. (See, for example, missing or multiple rulers, starts and ends of reigns in the same year, and rulers that appear several times with different territories.) Evan defining a monarch of England is tricky. Most of the monarchs in this dataset were around before England existed (it consisted of seven territories called the heptarchy). The data stops before John I (the bad guy from the Robin Hood stories) because he proclaimed himself King of Ireland, although some people consider monarchs up to Anne, five hundred years later, to be English Monarchs even though they ruled over Ireland, Wales and Scotland to varying degrees.

The heptarchy consisted of East Anglia, Essex, Kent, Mercia, Northumbria, Sussex and Wessex. Northubria was originally divided into Deria and Bernicia. There are also periods of Norse and Danish rule.

The dataset was compiled from Wikipedia and thus is not guaranteed to be error free. Please double check the data if you want to use it for something serious.

References

Richard Cotton richierocks@gmail.com compiled the dataset from various Wikipedia pages. http://en.wikipedia.org/wiki/Kings_of_england http://en.wikipedia.org/wiki/Kings_of_East_Anglia http://en.wikipedia.org/wiki/Kings_of_Essex http://en.wikipedia.org/wiki/Kings_of_Kent http://en.wikipedia.org/wiki/Kings_of_Mercia http://en.wikipedia.org/wiki/Kings_of_Northumbria http://en.wikipedia.org/wiki/Kings_of_Sussex http://en.wikipedia.org/wiki/Kings_of_Wessex

Convert filenames to match those in the book

Description

Some filenames have been altered in order to comply with portability requirements on CRAN. This function converts the filenames between the CRAN forms and the book forms.

Usage

  fix_filenames(x = c("tobook", "tocran"),
    dir = system.file("extdata", packages = "learningr"))

Arguments

Value

A logical vector of length 4, TRUE for each file whose name was changed.

Examples

## Not run: 
#To convert the files to the book form, use:
fix_filenames("tobook")
#The files were converted to CRAN form using:
fix_filenames("tocran", "learningr/inst/extdata")

## End(Not run)

Gonorrhoea

Description

Rates of gonorrhoea infection in the US by year, with contextual information about age, ethnicity and gender.

Format

A data frame with the following columns.

Year

Year that infected people visited the clinic.

Age.Group

Age group of person infected.

Ethnicity

Ethnicity of person infected.

Gender

Gender of person infected.

Rate

Number of infections per 100000 people.

References

Compiled by Richard Cotton richierocks@gmail.com from http://www.cdc.gov/std/stats11/tables/22b.htm

Hafu

Description

Half-caste manga characters.

Format

Both data frames have the following columns.

Year

Integer year that the manga was made.

Series

Name of series.

Character

Name of character.

Gender

Gender of character.

Father

Nationality of character's father.

Mother

Nationality of character's mother.

Eyes

Character's eye colour.

Hair

Character's hair colour.

Notes

Notes on data collection or ambiguity.

hafu2 has these additional columns.

FathersNationalityIsUncertain

MothersNationalityIsUncertain

Note

The dataset was kindly provided by Gwern Branwen.

hafu2 is a lightly cleaned up version of hafu.

Gwern's notes: The following list includes manga, light novel, anime, and video game characters (there being little point in keeping the mediums separate). It also includes characters who are not hafu themselves but a quarter-foreign inasmuch as they imply a hafu at some point. Characters are treated separately even if they are in the same work (eg. siblings). Classification is based on in-universe or out-of- universe information, since appearance can be highly misleading in anime (blue eyes may indicate heroic status, rather than being Caucasian; hair color may be chosen for contrast against other characters or signal stereotypes like red hair indicating a fiery personality), and different groups will identify the same anime character as belonging to their own race (Lu 2009), perhaps due to minimalistic drawings intended to save money or enable viewers to project themselves onto a character.

References

The dataset was compiled by Gwern Branwen gwern0@gmail.com. The original is available from http://www.gwern.net/hafu#list.

Dumb hypotenuse function.

Description

Calculate the (Pythagorean) hypotenuse of two numeric vectors using the obvious algorithm.

Usage

  hypotenuse(x, y)

Arguments

Value

A numeric vector of the hyptenuse of the inputs.

Note

This algorithm fails when the inputs are very large of very small, making it unsuitable for real-world use.

References

Cleve Moler (MATLAB creator and discoverer of the Moler-Morrison algorithm for calculating hypotenuses) discusses the pro and cons of several algorithms here. http://blogs.mathworks.com/cleve/2012/07/30/pythagorean-addition

See Also

hypot

Examples

hypotenuse(5, 12)           #okay
hypotenuse(1e-300, 1e-300)  #fails
hypotenuse(1e300, 1e300)    #fails

Obama vs. McCain

Description

State-by-state voting information in the 2008 US presidential election, along with contextual information on income, unemployment, ethnicity and religion.

Format

A data frame with 52 observations (one for each US state) and the following columns.

State

The name of the US state.

Region

The US Federal region.

Obama

Percentage of voters who voted for Barack Obama in the 2008 presidential election.

McCain

Percentage of voters who voted for John McCain in the 2008 presidential election.

Turnout

Percentage of people who voted in the 2008 presidential election.

Unemployment

Percentage of people who are unemployed.

Income

Mean annual income in US dollars.

Population

Number of people living in the state.

Catholic

Percentage of people identifying as Catholic.

Protestant

Percentage of people identifying as Protestant.

Other

Percentage of people identifying as religious, but not Catholic or Protestant.

Non.religious

Percentage of people identifying as non-religious.

Black

Percentage of people identifying as black.

Latino

Percentage of people identifying as Latino.

Urbanization

Percentage of people living in an urban area.

Note

Religious identification data are not available for Alaska and Hawaii. The totals of these columns is generally less than 100, since some people didn't give an answer. The District of Columbia is included, even though it isn't a state. The dataset is not guaranteed to be error free. Please double check the data if you want to use it for something serious.

References

This dataset was kindly compiled and provided by Edwin Thoen edwinthoen@hotmail.com.

The voting information came from http://www.uselectionatlas.org/, extracted on 2011-12-09.

The ethnicity, income and urbanisation information came from http://quickfacts.census.gov, extracted on 2011-12-09.

The unemployment information came from http://data.bls.gov/timeseries/LNS14000000, extracted 2011-12-09.

The religious information came from Table 12 of the American Religious Identification Survey 2008. http://commons.trincoll.edu/aris/files/2011/08/ARIS_Report_2008.pdf.