LearnEDAfunctions

This package includes special functions and datasets for a course on Exploratory Data Analysis based on John Tukey's EDA text.

Installation

It can be installed from Jim Albert's Github site:

library(remotes)
install_github("bayesball/LearnEDAfunctions")

The only prerequisite packages are the dplyr and ggplot2 packages.

Loading

The LearnEDAfunctions package is loaded by the library() function:

library(LearnEDAfunctions)

Other EDA Functions

Also there are several functions from other packages used in the EDA course. One will need to install these packages to use this functions.

• stem.leaf() from the aplpack package
• rootogram() from the vcd package

Datasets in LearnEDAfunctions

• act.scores.06.07 ACT scores of states in the US.
• baseball.attendance Attendance at baseball teams.
• batting.history Yearly Batting Statistics for Major League Baseball
• beatles Lengths of songs from Beatles' albums.
• boston.marathon Boston Marathon completion times of women of different ages.
• boston.marathon.wtimes Boston marathon winning times
• braves.attendance Attendance at games of a baseball team
• car.measurements Car measurements -church.2way Church attendance as a two-way table.
• church.tseries Time series of worship attendance
• college.ratings Ratings of National Universities in the U.S.
• farms Number of farms in the states of the U.S.
• fit.gaussian Fitting a Gaussian curve to binned data
• football American football scores
• gestation.periods Gestation periods for different animals.
• grandma.19.40 Grandmas Marathon completion times
• heaviest.fish Fish world record catches
• home.prices Home sales prices in the U.S.
• homeruns.2000 Team home run numbers for different seasons.
• homeruns.61 Home run counts in 1961
• immigrants Immigrant counts to US.
• island.areas Areas of islands from different continents.
• lake Lake measurements
• mortality.rates Infant mortality rates of countries
• olympics.run Olympics running times.
• olympics.speed.skating Olympics speed skating times.
• olympics.swim Olympics swimming times.
• pitching.history Yearly Pitching Statistics for Major League Baseball
• pop.change Population change for all states in the U.S.
• pop.densities Population densities of states for different years.
• pop.england England population
• rent.prices Rent prices in different cities.
• salaries Salaries of different professions in different cities.
• snowfall Snowfall amounts of two cities
• studentdata Student dataset
• temperatures Temperatures for different cities.
• tukey.24a Tukey straightening exercise 24a
• tukey.24b Tukey straightening exercise 24b
• tukey.26a Tukey straightening exercise 24a
• tukey.26b Tukey straightening exercise 24b
• tukey.26c Tukey straightening exercise 24c
• us.pop Population of United States

Functions in LearnEDAfunctions

fit.gaussian() function

Fits a Gaussian curve to binned data.

data <- rt(200,df=5)
bins <- pretty(range(data))
g.mean <- 0
g.sd <- 1
fit.gaussian(data, bins, g.mean, g.sd)

half.slope.ratio() function

Half-slope ratio

sx <- c(10,30,50)
sy <- c(5,8,20)
half.slope.ratio(sx, sy, 1, 1)
half.slope.ratio(sx, sy, -0.5, -0.5)

han.R() function

Hanning a sequence.

plot(WWWusage)
plot(smooth(WWWusage, kind="3RSS"))
plot(han(smooth(WWWusage, kind="3RSS")))

hinkley() function

Hinkley's quick method.

raw <- state.x77[, "Population"]
hinkley(raw)
logs <- log(raw)
hinkley(logs)

lval_plus() function

Finds outliers by group using Tukey's rule

lval_plus(beatles, time, album)

lval() function

Letter values

lval(rnorm(100))

mtrans() function

raw <- state.x77[,"Population"]
matched.roots <- mtrans(raw,0.5)
matched.logs <- mtrans(raw,0)
boxplot(data.frame(raw, matched.roots,
                   matched.logs))

plot2way() function

Plot of an additive fit.

temps <- matrix(data=c(50, 30, 35, 21, 38,
              73, 58, 65, 57, 63,
              88, 83, 89, 84, 86,
              73, 62, 68, 59, 66),
             nrow=5,ncol=4,
          dimnames=list(c("Atlanta",
              "Detroit", "Kansas City",
          "Minneapolis", "Philadelphia"),
          c("January", "April",
              "July", "October")))
fit <- medpolish(temps)
plot2way(fit$row + fit$overall, fit$col,
  dimnames(temps)[[1]],
  dimnames(temps)[[2]])

power.t() function

Power transformation.

power.t(c(3, 6, 5, 4, 7), 0.5)

rline() function

Computation of a resistant line.

df <- data.frame(x = 1:10,
          y = 3 * (1:10) + rnorm(10))
rline(y ~ x, df, iter=5)

spread_level_plot() function

Spread versus level plot.

spread_level_plot(beatles, time, album)

symplot() function

Symmetry plot.

symplot(rnorm(100))
# symmetry plot for exponential data
symplot(rexp(100))