Visualize univariate and bivariate distributions

TODO

• link to diagCategorical, diagScatter, diagMultivariate, diagAddElements, diagBounding

Install required packages

car, hexbin, beeswarm

wants <- c("car", "hexbin", "beeswarm")
has   <- wants %in% rownames(installed.packages())
if(any(!has)) install.packages(wants[!has])

Histograms

Histogram with absolute class frequencies

set.seed(123)
x <- rnorm(200, 175, 10)
hist(x, xlab="x", ylab="N", breaks="FD")

Add individual values and normal probability density function

hist(x, freq=FALSE, xlab="x", ylab="relative frequency",
     breaks="FD", main="Histogram und normal PDF")
rug(jitter(x))
curve(dnorm(x, mean(x), sd(x)), lwd=2, col="blue", add=TRUE)

Add estimated probability density function

hist(x, freq=FALSE, xlab="x", breaks="FD",
     main="Histogram and density estimate")
lines(density(x), lwd=2, col="blue")
rug(jitter(x))

To compare the histograms from two groups, see histbackback() from package Hmisc.

Stem and leaf plot

y <- rnorm(100, mean=175, sd=7)
stem(y)

  The decimal point is 1 digit(s) to the right of the |

  15 | 669
  16 | 134
  16 | 5566777789
  17 | 0011112222233333334444444
  17 | 5555566666677777788888888999
  18 | 0000000001111233334444
  18 | 55667779
  19 | 2

Boxplot

Nj <- 40
P  <- 3
DV <- rnorm(P*Nj, mean=100, sd=15)
IV <- gl(P, Nj, labels=c("Control", "Group A", "Group B"))
Mj <- tapply(DV, IV, FUN=mean)

boxplot(DV ~ IV, ylab="Score", col=c("red", "blue", "green"),
        main="Boxplot of scores in 3 groups")
points(1:P, Mj, pch=16, cex=2)

DVrange <- round(range(DV), digits=-1)
library(beeswarm)
boxplot(DV ~ IV,
        ylab="Score", col=c("red", "blue", "green"),
        main="Boxplots und Scores in 3 Gruppen",
        outline=FALSE, ylim=DVrange)

beeswarm(DV ~ IV, add=TRUE, pch=16, col="#00000077")

Dotchart

Nj  <- 5
DV1 <- rnorm(Nj, 20, 2)
DV2 <- rnorm(Nj, 25, 2)
DV  <- c(DV1, DV2)
IV  <- gl(2, Nj)
Mj  <- tapply(DV, IV, FUN=mean)

dotchart(DV, gdata=Mj, pch=16, color=rep(c("red", "blue"), each=Nj),
         gcolor="black", labels=rep(LETTERS[1:Nj], 2), groups=IV,
         xlab="AV", ylab="group",
         main="individual results and means from 2 groups")

Stripchart

Nj   <- 25
P    <- 4
dice <- sample(1:6, P*Nj, replace=TRUE)
IV   <- gl(P, Nj)

stripchart(dice ~ IV, xlab="Result", ylab="group", pch=1, col="blue",
           main="Dice results: 4 groups", sub="jitter-method", method="jitter")

stripchart(dice ~ IV, xlab="Result", ylab="group", pch=16, col="red",
           main="Dice results: 4 groups", sub="stack-method", method="stack")

QQ-plot

DV1 <- rnorm(200)
DV2 <- rf(200, df1=3, df2=15)
qqplot(DV1, DV2, xlab="quantile N(0, 1)", ylab="quantile F(3, 15)",
       main="Comparison of quantiles from N(0, 1) and F(3, 15)")

height <- rnorm(100, mean=175, sd=7)
qqnorm(height)
qqline(height, col="red", lwd=2)

Empirical cumulative distribution function

vec <- round(rnorm(10), 1)
Fn  <- ecdf(vec)
plot(Fn, main="Empirical cumulative distribution function")
curve(pnorm, add=TRUE, col="gray", lwd=2)

Joint distribution of two variables in separate groups

Simulate data

N  <- 200
P  <- 2
x  <- rnorm(N, 100, 15)
y  <- 0.5*x + rnorm(N, 0, 10)
IV <- gl(P, N/P, labels=LETTERS[1:P])

Identify group membership by plot symbol and color

plot(x, y, pch=c(4, 16)[unclass(IV)], lwd=2,
     col=c("black", "blue")[unclass(IV)],
     main="Joint distribution per group")
legend(x="topleft", legend=c("group A", "group B"),
       pch=c(4, 16), col=c("black", "blue"))

Add distribution ellipse

Pooled groups

library(car)
dataEllipse(x, y, xlab="x", ylab="y", asp=1, levels=0.5, lwd=2, center.pch=16,
            col="blue", main="Joint distribution of two variables")
legend(x="bottomright", legend=c("Data", "centroid", "distribution ellipse"),
       pch=c(1, 16, NA), lty=c(NA, NA, 1), col=c("black", "blue", "blue"))

Joint distribution of two variables with many observations

Using transparency

N  <- 5000
xx <- rnorm(N, 100, 15)
yy <- 0.4*xx + rnorm(N, 0, 10)
plot(xx, yy, pch=16, col=rgb(0, 0, 1, 0.3))

Smooth scatter plot

Based on a 2-D kernel density estimate

smoothScatter(xx, yy, bandwidth=4)

Hexagonal 2-D binning

library(hexbin)
res <- hexbin(xx, yy, xbins=20)
plot(res)

summary(res)

'hexbin' object from call: hexbin(x = xx, y = yy, xbins = 20) 
n = 5000  points in nc = 213  hexagon cells in grid dimensions  26 by 21 
      cell           count             xcm              ycm        
 Min.   :  9.0   Min.   :  1.00   Min.   : 44.83   Min.   :-5.868  
 1st Qu.:157.0   1st Qu.:  2.00   1st Qu.: 81.12   1st Qu.:23.699  
 Median :241.0   Median :  8.00   Median :101.32   Median :40.057  
 Mean   :242.4   Mean   : 23.47   Mean   :101.14   Mean   :40.226  
 3rd Qu.:325.0   3rd Qu.: 31.00   3rd Qu.:120.87   3rd Qu.:56.396  
 Max.   :499.0   Max.   :134.00   Max.   :157.78   Max.   :90.498  

Detach (automatically) loaded packages (if possible)

try(detach(package:car))
try(detach(package:carData))
try(detach(package:hexbin))
try(detach(package:beeswarm))

Get the article source from GitHub

R markdown - markdown - R code - all posts