ggplot2 Version of Figures in “Lattice: Multivariate Data Visualization with R” (Part 3)

July 2, 2009

tags: boxplot, density, ecdf, ggplot2, histogram, jitter, lattice, qq, R, violin

This is the third post in a series attempting to recreate the figures in Lattice: Multivariate Data Visualization with R (R code) with ggplot2.

Chapter 3 – Visualizing Univariate Distributions

Topics covered:

Kernel Density Plot, Histogram
Theoretical Q-Q plot, Empirical CDF plot
Two-sample Q-Q plot
Comparative Box and Whisker plots, Violin plots
Comparative Strip charts
Discrete distributions

Figure 3.1

> library(lattice)
> library(ggplot2)
> data(Oats, package = "MEMSS")

lattice

> pl <- densityplot(~eruptions, data = faithful)
> print(pl)

ggplot2

> p <- ggplot(faithful, aes(eruptions))
> pg <- p + stat_density(geom = "path", position = "identity") +
+     geom_point(aes(y = 0.05), position = position_jitter(height = 0.005),
+         alpha = 0.25)
> print(pg)

Note

y = 0.05 specifies the position of jitter on y-axis.

Figure 3.2

lattice

> pl <- densityplot(~eruptions, data = faithful, kernel = "rect",
+     bw = 0.2, plot.points = "rug", n = 200)
> print(pl)

ggplot2

> pg <- p + stat_density(geom = "path", kernel = "rect",
+     position = "identity", bw = 0.2) + geom_rug()
> print(pg)

Figure 3.3

> library("latticeExtra")
> data(gvhd10)

lattice

> pl <- densityplot(~log(FSC.H) | Days, data = gvhd10,
+     plot.points = FALSE, ref = TRUE, layout = c(2, 4))
> print(pl)

ggplot2

> p <- ggplot(gvhd10, aes(log(FSC.H)))
> pg <- p + stat_density(geom = "path", position = "identity") +
+     facet_wrap(~Days, ncol = 2, as.table = FALSE)
> print(pg)

Note

as.table = FALSE changes the default orders of the facets.

Figure 3.4

lattice

> pl <- histogram(~log2(FSC.H) | Days, gvhd10, xlab = "log Forward Scatter",
+     type = "density", nint = 50, layout = c(2, 4))
> print(pl)

ggplot2

> pg <- p + geom_histogram(aes(y = ..density..), binwidth = diff(range(log2(gvhd10$FSC.H)))/50) +
+     facet_wrap(~Days, ncol = 2, as.table = FALSE) + xlab("log Forward Scatter")
> print(pg)

Note

ggplot2 uses binwidth by default, therefore the number of bins needs to be presented in terms of binwidth.

Figure 3.5

> data(Chem97, package = "mlmRev")

lattice

> pl <- qqmath(~gcsescore | factor(score), data = Chem97,
+     f.value = ppoints(100))
> print(pl)

ggplot2

> p <- ggplot(Chem97)
> pg <- p + geom_point(aes(sample = gcsescore), stat = "qq",
+     quantiles = ppoints(100)) + facet_wrap(~score)
> print(pg)

Figure 3.6

lattice

> pl <- qqmath(~gcsescore | gender, Chem97, groups = score,
+     aspect = "xy", f.value = ppoints(100), auto.key = list(space = "right"),
+     xlab = "Standard Normal Quantiles", ylab = "Average GCSE Score")
> print(pl)

ggplot2

> pg <- p + geom_point(aes(sample = gcsescore, colour = factor(score)),
+     stat = "qq", quantiles = ppoints(100)) + facet_grid(~gender) +
+     opts(aspect.ratio = 1) + scale_x_continuous("Standard Normal Quantiles") +
+     scale_y_continuous("Average GCSE Score")
> print(pg)

Figure 3.7

lattice

> Chem97.mod <- transform(Chem97, gcsescore.trans = gcsescore^2.34)

> pl <- qqmath(~gcsescore.trans | gender, Chem97.mod, groups = score,
+     f.value = ppoints(100), aspect = "xy", auto.key = list(space = "right",
+         title = "score"), xlab = "Standard Normal Quantiles",
+     ylab = "Transformed GCSE Score")
> print(pl)

ggplot2

> pg <- p + geom_point(aes(sample = gcsescore^2.34, colour = factor(score)),
+     stat = "qq", quantiles = ppoints(100)) + facet_grid(~gender) +
+     opts(aspect.ratio = 1) + scale_x_continuous("Standard Normal Quantiles") +
+     scale_y_continuous("Transformed GCSE Score")
> print(pg)

Figure 3.8

> library("latticeExtra")

lattice

> pl <- ecdfplot(~gcsescore | factor(score), data = Chem97,
+     groups = gender, auto.key = list(columns = 2), subset = gcsescore >
+         0, xlab = "Average GCSE Score")
> print(pl)

ggplot2

> Chem97.ecdf <- ddply(Chem97, .(score, gender), transform,
+     ecdf = ecdf(gcsescore)(gcsescore))

> p <- ggplot(Chem97.ecdf, aes(gcsescore, ecdf, colour = gender))
> pg <- p + geom_step(subset = .(gcsescore > 0)) + facet_wrap(~score,
+     as.table = F) + xlab("Average GCSE Score") + ylab("Empirical CDF")
> print(pg)

Figure 3.9

lattice

> pl <- qqmath(~gcsescore | factor(score), data = Chem97,
+     groups = gender, auto.key = list(points = FALSE,
+         lines = TRUE, columns = 2), subset = gcsescore >
+         0, type = "l", distribution = qunif, prepanel = prepanel.qqmathline,
+     aspect = "xy", xlab = "Standard Normal Quantiles",
+     ylab = "Average GCSE Score")
> print(pl)

ggplot2

> p <- ggplot(Chem97, aes(sample = gcsescore, colour = gender))
> pg <- p + geom_path(subset = .(gcsescore > 0), stat = "qq",
+     distribution = qunif) + facet_grid(~score) + scale_x_continuous("Standard Normal Quantiles") +
+     scale_y_continuous("Average GCSE Score")
> print(pg)

Figure 3.10

lattice

> pl <- qq(gender ~ gcsescore | factor(score), Chem97,
+     f.value = ppoints(100), aspect = 1)
> print(pl)

ggplot2

> q <- function(x, probs = ppoints(100)) {
+     data.frame(q = probs, value = quantile(x, probs))
+ }
> Chem97.q <- ddply(Chem97, c("gender", "score"), function(df) q(df$gcsescore))
> Chem97.df <- recast(Chem97.q, score + q ~ gender, id.var = 1:3)

> pg <- ggplot(Chem97.df) + geom_point(aes(M, F)) + geom_abline() +
+     facet_wrap(~score) + coord_equal()
> print(pg)

Figure 3.11

lattice

> pl <- bwplot(factor(score) ~ gcsescore | gender, data = Chem97,
+     xlab = "Average GCSE Score")
> print(pl)

ggplot2

> pg <- ggplot(Chem97, aes(factor(score), gcsescore)) +
+     geom_boxplot() + coord_flip() + ylab("Average GCSE score") +
+     facet_wrap(~gender)
> print(pg)

Figure 3.12

lattice

> pl <- bwplot(gcsescore^2.34 ~ gender | factor(score),
+     Chem97, varwidth = TRUE, layout = c(6, 1), ylab = "Transformed GCSE score")
> print(pl)

ggplot2

> p <- ggplot(Chem97, aes(factor(gender), gcsescore^2.34))
> pg <- p + geom_boxplot() + facet_grid(~score) + ylab("Transformed GCSE score")
> print(pg)

Figure 3.13

lattice

> pl <- bwplot(Days ~ log(FSC.H), data = gvhd10, xlab = "log(Forward Scatter)",
+     ylab = "Days Past Transplant")
> print(pl)

ggplot2

> p <- ggplot(gvhd10, aes(factor(Days), log(FSC.H)))
> pg <- p + geom_boxplot() + coord_flip() + labs(y = "log(Forward Scatter)",
+     x = "Days Past Transplant")
> print(pg)

Figure 3.14

lattice

> pl <- bwplot(Days ~ log(FSC.H), gvhd10, panel = panel.violin,
+     box.ratio = 3, xlab = "log(Forward Scatter)", ylab = "Days Past Transplant")
> print(pl)

ggplot2

> p <- ggplot(gvhd10, aes(log(FSC.H), Days))
> pg <- p + geom_ribbon(aes(ymax = ..density.., ymin = -..density..),
+     stat = "density") + facet_grid(Days ~ ., as.table = F,
+     scales = "free_y") + labs(x = "log(Forward Scatter)",
+     y = "Days Past Transplant")
> print(pg)

Figure 3.15

lattice

> pl <- stripplot(factor(mag) ~ depth, quakes)
> print(pl)

ggplot2

> pg <- ggplot(quakes) + geom_point(aes(depth, mag), shape = 1)
> print(pg)

Figure 3.16

lattice

> pl <- stripplot(depth ~ factor(mag), quakes, jitter.data = TRUE,
+     alpha = 0.6, xlab = "Magnitude (Richter)", ylab = "Depth (km)")
> print(pl)

ggplot2

> p <- ggplot(quakes, aes(factor(mag), depth))
> pg <- p + geom_point(position = position_jitter(width = 0.15),
+     alpha = 0.6, shape = 1) + theme_bw() + xlab("Magnitude (Richter)") +
+     ylab("Depth (km)")
> print(pg)

Figure 3.17

lattice

> pl <- stripplot(sqrt(abs(residuals(lm(yield ~ variety +
+     year + site)))) ~ site, data = barley, groups = year,
+     jitter.data = TRUE, auto.key = list(points = TRUE,
+         lines = TRUE, columns = 2), type = c("p", "a"),
+     fun = median, ylab = expression(abs("Residual Barley Yield")^{
+         1/2
+     }))
> print(pl)

ggplot2

> p <- ggplot(barley, aes(site, sqrt(abs(residuals(lm(yield ~
+     variety + year + site)))), colour = year, group = year))
> pg <- p + geom_jitter(position = position_jitter(width = 0.2)) +
+     geom_line(stat = "summary", fun.y = "mean") + labs(x = "",
+     y = expression(abs("Residual Barley Yield")^{
+         1/2
+     }))
> print(pg)

17 Comments leave one →

Hadley Wickham permalink

July 2, 2009 4:57 pm

Great again 🙂 You can simplify the code to produce the ecdf dataset a little:

Chem97.ecdf <- ddply(Chem97, .(score, gender), transform,
ecdf = ecdf(gcsescore)(gcsescore)
)

Reply
- learnr permalink*
  
  August 3, 2009 9:49 am
  
  Thanks. I have updated the post.
  
  Reply
JD Long permalink

July 2, 2009 8:45 pm

I’m really enjoying your tutorial series. You’ve sunk a lot of time into these!

Drop me an email when you get a chance. A group of R folks have been organizing some community online activities and I would like your input on a few things.

-JD

Reply
Ben Bolker permalink

July 6, 2009 6:55 am

truly tiny, but: any way to get rid of the extra labels on the RHS in figure 3.14?

very nice (and impressive to define the violin plot on the fly like that!)

Reply
- learnr permalink*
  
  July 7, 2009 2:33 am
  
  I would rather get rid of the the labels on the left.
  However, to get rid of the labels on the right the following code should do the trick (hides them almost completely):
  pg + opts(strip.background = theme_rect(fill = NA, colour = NA), strip.text.y = theme_text(size = 0))
  
  Reply
asmus permalink

July 29, 2009 1:21 pm

Hi,

this is just what I needed to find. Thanks a lot for the very fine examples!

Reply
Dennis Murphy permalink

December 8, 2009 3:47 pm

Hi – very nice job! One question: shouldn’t the second piece of ggplot2 code for Figure 3.12 be

pg <- p + geom_boxplot() + facet_grid(~score) + ylab("Transformed GCSE score")

?

Reply
- learnr permalink*
  
  December 8, 2009 5:35 pm
  
  Thank you for spotting the typo. It is fixed now.
  
  Reply
dengyishuo permalink

January 1, 2010 4:39 am

Your blog is very nice! Good luck to you.

Reply
volatile smile permalink

November 2, 2010 10:38 am

WOW. This is a _great_ comparison.

But just like excelcharts.com’s fools in his office that like a multi-coloured bubble chart better than an unfilled grey one … I can’t tell which I like better!

I’m not even sure whether the ggplot axes are Tufte-compliant. I mean, you’re erasing, but you’re also using more ink. Does “erasing the paper” count as reducing ink? Or adding colour?

So I guess the main issue is, I can’t tell if the grid lines are distracting or add clarity.

Reply
Evgenia Martynova permalink

December 29, 2010 5:52 pm

Hi,

First of all thank you for the tutorial. But when I am trying the above examples, I’ve got the following annoying error:
gender))

> p p
Ошибка: No layers in plot
Could you please help what is wrong? R version (2.12.0) for Windows

Reply
- Evgenia Martynova permalink
  
  December 29, 2010 6:34 pm
  
  Just upgraded R to 2.12.1 – everything is working now. Thanks a again!!!
  
  Reply

	S.Stender on ggplot2: Labelling Data Series…
	ggplot séparer la lé… on ggplot2: Two Or More Plots Sha…
	9 Useful R Data Visu… on ggplot2 Version of Figures in…
	Mandar on Data Manipulation in R to Crea…
	Mandar on Data Manipulation in R to Crea…

Finding my way around R

ggplot2 Version of Figures in “Lattice: Multivariate Data Visualization with R” (Part 3)

Chapter 3 – Visualizing Univariate Distributions

Figure 3.1

Figure 3.2

Figure 3.3

Figure 3.4

Figure 3.5

Figure 3.6

Figure 3.7

Figure 3.8

Figure 3.9

Figure 3.10

Figure 3.11

Figure 3.12

Figure 3.13

Figure 3.14

Figure 3.15

Figure 3.16

Figure 3.17

Trackbacks

Leave a Reply Cancel reply

Subscribe

Recent Posts

Recent Comments

Archives

Finding my way around R

ggplot2 Version of Figures in “Lattice: Multivariate Data Visualization with R” (Part 3)

Chapter 3 – Visualizing Univariate Distributions

Figure 3.1

Figure 3.2

Figure 3.3

Figure 3.4

Figure 3.5

Figure 3.6

Figure 3.7

Figure 3.8

Figure 3.9

Figure 3.10

Figure 3.11

Figure 3.12

Figure 3.13

Figure 3.14

Figure 3.15

Figure 3.16

Figure 3.17

Share this:

Like this:

Related

Trackbacks

Leave a Reply Cancel reply

Subscribe

Recent Posts

Recent Comments

Archives