R Notebook

Data Visualization in ggplot

Penn SAS Data Driven Discovery Summer Hangouts 2023

Brynn Sherman (brynns@sas.upenn.edu)

DDDI postdoctoral fellow | Department of Psychology

Load in the relevant packages

library(tidyverse)
library(datasets)

Check out the Iris dataset

?iris

iris

Plot sepal length and sepal width against one another

ggplot(data = iris, aes(x = Sepal.Length, y = Sepal.Width)) + geom_point()

Add a line of best fit

ggplot(data = iris, aes(x = Sepal.Length, y = Sepal.Width)) + geom_point() + geom_smooth(method = "lm")

`geom_smooth()` using formula 'y ~ x'

Let’s separate out based on species

First, by color:

ggplot(data = iris, aes(x = Sepal.Length, y = Sepal.Width, color = Species)) + geom_point() + geom_smooth(method = "lm")

`geom_smooth()` using formula 'y ~ x'

Next, by facets

ggplot(data = iris, aes(x = Sepal.Length, y = Sepal.Width)) + geom_point() + geom_smooth(method = "lm") + facet_wrap(~Species)

`geom_smooth()` using formula 'y ~ x'

Note that by default, R has the scale of the three subplots as the same. How can we change that?

ggplot(data = iris, aes(x = Sepal.Length, y = Sepal.Width)) + geom_point() + geom_smooth(method = "lm") + facet_wrap(~Species, scales = "free")

`geom_smooth()` using formula 'y ~ x'

We can also change the shape of the points associated with the three species

ggplot(data = iris, aes(x = Sepal.Length, y = Sepal.Width, shape = Species)) + geom_point() + geom_smooth(method = "lm")

`geom_smooth()` using formula 'y ~ x'

ggplot(data = iris, aes(x = Sepal.Length, y = Sepal.Width, shape = Species)) + geom_point(alpha = .25) + geom_smooth(method = "lm")

`geom_smooth()` using formula 'y ~ x'

Plotting the average petal length for each species

groupedData = group_by(iris, Species) %>% summarise(meanPetalLength = mean(Petal.Length))
groupedData

ggplot(data = groupedData,aes(x = Species, y = meanPetalLength)) + geom_bar(stat = "identity")

The bars represent the means, which isn’t the most useful. Ideally, we’d also like a measure of variance.

One way to do this is to add error bars (in this case, standard error of the mean)

groupedData = group_by(iris, Species) %>% summarise(meanPetalLength = mean(Petal.Length),sdPetalLength = sd(Petal.Length)/sqrt(n()))
groupedData

ggplot(data = groupedData,aes(x = Species, y = meanPetalLength)) + geom_bar(stat = "identity") + geom_errorbar(data = groupedData,aes(x = Species, ymin = meanPetalLength - sdPetalLength, ymax = meanPetalLength + sdPetalLength))

ggplot(data = groupedData,aes(x = Species, y = meanPetalLength)) + geom_bar(stat = "identity") + geom_errorbar(data = groupedData,aes(x = Species, ymin = meanPetalLength - sdPetalLength, ymax = meanPetalLength + sdPetalLength),width=.1)

Aside: Another way to plot the mean more directly (without creating a new dataframe)

ggplot(data = iris,aes(x = Species, y = Petal.Length)) + geom_bar(stat = "summary", fun.y = "mean")

Warning: Ignoring unknown parameters: fun.y
No summary function supplied, defaulting to `mean_se()`

What if we want to get a better sense of the distribution of petal lengths for each species (not just the mean/sd)?

Histograms

ggplot(data = iris, aes(x = Petal.Length,fill = Species)) + geom_histogram()

`stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

ggplot(data = iris, aes(x = Petal.Length)) + geom_histogram() + facet_wrap(~Species)

`stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

Boxplot

ggplot(data = iris, aes(x = Species, y = Petal.Length)) + geom_boxplot() 

Dotplot

ggplot(data = iris, aes(x = Species, y = Petal.Length)) + geom_dotplot(binaxis = "y",stackdir = "center")

Bin width defaults to 1/30 of the range of the data. Pick better value with `binwidth`.

ggplot(data = iris, aes(x = Species, y = Petal.Length)) + geom_dotplot(binaxis = "y",stackdir = "center",dotsize=.5)

Bin width defaults to 1/30 of the range of the data. Pick better value with `binwidth`.

Violin plot

ggplot(data = iris, aes(x = Species, y = Petal.Length)) + geom_violin()

Violin + boxplot

ggplot(data = iris, aes(x = Species, y = Petal.Length)) + geom_violin() + geom_dotplot(binaxis = "y",stackdir = "center",dotsize=.5,alpha=.5)

Bin width defaults to 1/30 of the range of the data. Pick better value with `binwidth`.

Plotting the individual data points on the mean bars

groupedData

ggplot(data = groupedData,aes(x = Species, y = meanPetalLength)) + geom_bar(stat = "identity") + geom_dotplot(data = iris, aes(x = Species, y = Petal.Length), binaxis = "y", stackdir = "center",dotsize = .5)

Bin width defaults to 1/30 of the range of the data. Pick better value with `binwidth`.

Now let’s spruce up the graph

First fix the y axis label

ggplot(data = groupedData,aes(x = Species, y = meanPetalLength)) + geom_bar(stat = "identity") + geom_dotplot(data = iris, aes(x = Species, y = Petal.Length), binaxis = "y", stackdir = "center",dotsize = .5) + ylab('Petal Length')

Bin width defaults to 1/30 of the range of the data. Pick better value with `binwidth`.

Change the theme

ggplot(data = groupedData,aes(x = Species, y = meanPetalLength)) + geom_bar(stat = "identity") + geom_dotplot(data = iris, aes(x = Species, y = Petal.Length), binaxis = "y", stackdir = "center",dotsize = .5) + ylab('Petal Length') + theme_classic()

Bin width defaults to 1/30 of the range of the data. Pick better value with `binwidth`.

Color by species

ggplot(data = groupedData,aes(x = Species, y = meanPetalLength, fill = Species)) + geom_bar(stat = "identity") + geom_dotplot(data = iris, aes(x = Species, y = Petal.Length), binaxis = "y", stackdir = "center",dotsize = .5) + ylab('Petal Length') + theme_classic()

Bin width defaults to 1/30 of the range of the data. Pick better value with `binwidth`.

Change the color scheme

library(RColorBrewer)

ggplot(data = groupedData,aes(x = Species, y = meanPetalLength, fill = Species)) + geom_bar(stat = "identity") + geom_dotplot(data = iris, aes(x = Species, y = Petal.Length), binaxis = "y", stackdir = "center",dotsize = .5) + ylab('Petal Length') + theme_classic() + scale_fill_brewer(palette = "Accent")

Bin width defaults to 1/30 of the range of the data. Pick better value with `binwidth`.

Change the font, font size

ggplot(data = groupedData,aes(x = Species, y = meanPetalLength, fill = Species)) + geom_bar(stat = "identity") + geom_dotplot(data = iris, aes(x = Species, y = Petal.Length), binaxis = "y", stackdir = "center",dotsize = .5) + ylab('Petal Length') + theme_classic() + scale_fill_brewer(palette = "Accent") + theme(text = element_text(size = 20,family = "mono"))

Bin width defaults to 1/30 of the range of the data. Pick better value with `binwidth`.

# save the plot
ggsave('petal_means.pdf',width=5,height=5)

Bin width defaults to 1/30 of the range of the data. Pick better value with `binwidth`.

Change the size of the plot so that it’s not cut off when it saves

ggplot(data = groupedData,aes(x = Species, y = meanPetalLength, fill = Species)) + geom_bar(stat = "identity") + geom_dotplot(data = iris, aes(x = Species, y = Petal.Length), binaxis = "y", stackdir = "center",dotsize = .5) + ylab('Petal Length') + theme_classic() + scale_fill_brewer(palette = "Accent") + theme(text = element_text(size = 20,family = "mono"))

Bin width defaults to 1/30 of the range of the data. Pick better value with `binwidth`.

# save the plot
ggsave('petal_means_wide.pdf',width=7,height=5)

Bin width defaults to 1/30 of the range of the data. Pick better value with `binwidth`.

Alternatively: is a legend really necessary in this plot?

ggplot(data = groupedData,aes(x = Species, y = meanPetalLength, fill = Species)) + geom_bar(stat = "identity") + geom_dotplot(data = iris, aes(x = Species, y = Petal.Length), binaxis = "y", stackdir = "center",dotsize = .5) + ylab('Petal Length') + theme_classic() + scale_fill_brewer(palette = "Accent") + theme(text = element_text(size = 20,family = "mono")) + theme(legend.position = "none")

Bin width defaults to 1/30 of the range of the data. Pick better value with `binwidth`.

# save the plot
ggsave('petal_means_noLegend.pdf',width=5,height=5)

Bin width defaults to 1/30 of the range of the data. Pick better value with `binwidth`.

Load in a different dataset (mpg) to illustrate a few other specific instances that come up often

?mpg

mpg

How do highway and city mpg relate to one another, and is there an effect of the model year?

ggplot(data = mpg,aes(x = cty, y = hwy, color = cyl)) + geom_point() + geom_smooth(method = "lm")

`geom_smooth()` using formula 'y ~ x'

The above graph treats cylinder as a continuous variable (which is probably okay in this case). But what if it is discrete variable?

ggplot(data = mpg,aes(x = cty, y = hwy, color = as.factor(cyl))) + geom_point() + geom_smooth(method = "lm")

`geom_smooth()` using formula 'y ~ x'

# alternatively, could also change it within the dataframe:
# mpg$cyl = factor(mpg$cyl)

Changing it to a factor means that we get individual lines of best fit for each level of the variable. Here’s a work-around to avoid that

ggplot(data = mpg,aes(x = cty, y = hwy)) + geom_point(aes(color = as.factor(cyl))) + geom_smooth(method = "lm")

`geom_smooth()` using formula 'y ~ x'

mpg

What is the average highway mpg by year for each manufacturer?

groupedData = group_by(mpg, manufacturer, year) %>% summarise(meanMPG = mean(hwy),sdMPG = sd(hwy)/sqrt(n()))

`summarise()` has grouped output by 'manufacturer'. You can override using the `.groups` argument.

groupedData

For the sake of space, let’s only plot Audi, Chevy, Dodge, Ford, and Honda

filteredData = filter(groupedData,manufacturer %in% c("audi","chevrolet","dodge","ford","honda"))
filteredData

ggplot(data = filteredData,aes(x = manufacturer, y = meanMPG, fill = year)) + geom_bar(stat = "identity") + geom_errorbar(data = filteredData,aes(x = manufacturer, ymin = meanMPG - sdMPG, ymax = meanMPG + sdMPG),width=.5)

Unstack the bars

ggplot(data = filteredData,aes(x = manufacturer, y = meanMPG, fill = year)) + geom_bar(stat = "identity", position = "dodge") + geom_errorbar(data = filteredData,aes(x = manufacturer, ymin = meanMPG - sdMPG, ymax = meanMPG + sdMPG),width=.5, position = position_dodge(.9))

Why did that not work?

ggplot(data = filteredData,aes(x = manufacturer, y = meanMPG, fill = as.factor(year))) + geom_bar(stat = "identity", position = "dodge") + geom_errorbar(data = filteredData,aes(x = manufacturer, ymin = meanMPG - sdMPG, ymax = meanMPG + sdMPG),width=.5, position = position_dodge(.9))

Make generative art! Courtesy of https://r-graph-gallery.com/137-spring-shapes-data-art.html

set.seed(567)
ngroup=30
names=paste("G_",seq(1,ngroup),sep="")
DAT=data.frame()

for(i in seq(1:30)){
    data=data.frame( matrix(0, ngroup , 3))
    data[,1]=i
    data[,2]=sample(names, nrow(data))
    data[,3]=prop.table(sample( c(rep(0,100),c(1:ngroup)) ,nrow(data)))
    DAT=rbind(DAT,data)
    }
colnames(DAT)=c("Year","Group","Value")
DAT=DAT[order( DAT$Year, DAT$Group) , ]


coul = brewer.pal(12, "Paired") 
coul = colorRampPalette(coul)(ngroup)
coul=coul[sample(c(1:length(coul)) , size=length(coul) ) ]

ggplot(DAT, aes(x=Year, y=Value, fill=Group )) + 
    geom_area(alpha=1  )+
    theme_bw() +
    #scale_fill_brewer(colour="red", breaks=rev(levels(DAT$Group)))+
    scale_fill_manual(values = coul)+
     theme(
        text = element_blank(),
        line = element_blank(),
        title = element_blank(),
        legend.position="none",
        panel.border = element_blank(),
        panel.background = element_blank())
ggsave('generative_art.pdf',width=7,height=5)