normalizing ggplot2 densities with facet_wrap in R - r

I am making a series of density plots with geom_density from a dataframe, and showing it by condition using facet_wrap, as in:
ggplot(iris) + geom_density(aes(x=Sepal.Width, colour=Species, y=..count../sum(..count..))) + facet_wrap(~Species)
When I do this, the y-axis scale seems to not represent percent of each Species in a panel, but rather the percent of all the total datapoints across all species.
My question is: How can I make it so the ..count.. variable in geom_density refers to the count of items in each Species set of each panel, so that the panel for virginica has a y-axis corresponding to "Fraction of virginica data points"?
Also, is there a way to get ggplot2 to output the values it uses for ..count.. and sum(..count..) so that I can verify what numbers it is using?
edit: I misunderstood geom_density it looks like even for a single Species, ..count../sum(..count..) is not a percentage:
ggplot(iris[iris$Species == 'virginica',]) + geom_density(aes(x=Sepal.Width, colour=Species, y=..count../sum(..count..))) + facet_wrap(~Species)
so my revised question: how can I get the density plot to be the fraction of data in each bin? Do I have to use stat_density for this or geom_histogram? I just want the y-axis to be percentage / fraction of data points

Unfortunately, what you are asking ggplot2 to do is define separate y's for each facet, which it syntactically cannot do AFAIK.
So, in response to your mentioning in the comment thread that you "just want a histogram fundamentally", I would suggest instead using geom_histogram or, if you're partial to lines instead of bars, geom_freqpoly:
ggplot(iris, aes(Sepal.Width, ..count..)) +
geom_histogram(aes(colour=Species, fill=Species), binwidth=.2) +
geom_freqpoly(colour="black", binwidth=.2) +
facet_wrap(~Species)
**Note: geom_freqpoly works just as well in place of geom_histogram in my above example. I just added both in one plot for sake of efficiency.
Hope this helps.
EDIT: Alright, I managed to work out a quick-and-dirty way of getting what you want. It requires that you install and load plyr. Apologies in advance; this is likely not the most efficient way to do this in terms of RAM usage, but it works.
First, let's get iris out in the open (I use RStudio so I'm used to seeing all my objects in a window):
d <- iris
Now, we can use ddply to count the number of individuals belonging to each unique measurement of what will become your x-axis (here I used Sepal.Length instead of Sepal.Width, to give myself a bit more range, simply for seeing a bigger difference between groups when plotted).
new <- ddply(d, c("Species", "Sepal.Length"), summarize, count=length(Sepal.Length))
Note that ddply automatically sorts the output data.frame according to the quoted variables.
Then we can divvy up the data.frame into each of its unique conditions--in the case of iris, each of the three species (I'm sure there's a much smoother way to go about this, and if you're working with really large amounts of data it's not advisable to keep creating subsets of the same data.frame because you could max out your RAM)...
set <- new[which(new$Species%in%"setosa"),]
ver <- new[which(new$Species%in%"versicolor"),]
vgn <- new[which(new$Species%in%"virginica"),]
... and use ddply again to calculate proportions of individuals falling under each measurement, but separately for each species.
prop <- rbind(ddply(set, c("Species"), summarize, prop=set$count/sum(set$count)),
ddply(ver, c("Species"), summarize, prop=ver$count/sum(ver$count)),
ddply(vgn, c("Species"), summarize, prop=vgn$count/sum(vgn$count)))
Then we just put everything we need into one dataset and remove all the junk from our workspace.
new$prop <- prop$prop
rm(list=ls()[which(!ls()%in%c("new", "d"))])
And we can make our figure with facet-specific proportions on the y. Note that I'm now using geom_line since ddply has automatically ordered your data.frame.
ggplot(new, aes(Sepal.Length, prop)) +
geom_line(aes(colour=new$Species)) +
facet_wrap(~Species)
# let's check our work. each should equal 50
sum(new$count[which(new$Species%in%"setosa")])
sum(new$count[which(new$Species%in%"versicolor")])
sum(new$count[which(new$Species%in%"versicolor")])
#... and each of these should equal 1
sum(new$prop[which(new$Species%in%"setosa")])
sum(new$prop[which(new$Species%in%"versicolor")])
sum(new$prop[which(new$Species%in%"versicolor")])

Maybe using table() and barplot() you might be able to get what you need. I'm still not sure if this is what you are after...
barplot(table(iris[iris$Species == 'virginica',1]))
With ggplot2
tb <- table(iris[iris$Species == 'virginica',1])
tb <- as.data.frame(tb)
ggplot(tb, aes(x=Var1, y=Freq)) + geom_bar()

Passing the argument scales='free_y' to facet_wrap() should do the trick.

Related

Histogram: Combine continuous and discrete values in ggplot2

I have a set of times that I would like to plot on a histogram.
Toy example:
df <- data.frame(time = c(1,2,2,3,4,5,5,5,6,7,7,7,9,9, ">10"))
The problem is that one value is ">10" and refers to the number of times that more than 10 seconds were observed. The other time points are all numbers referring to the actual time. Now, I would like to create a histogram that treats all numbers as numeric and combines them in bins when appropriate, while plotting the counts of the ">10" at the side of the distribution, but not in a separate plot. I have tried to call geom_histogram twice, once with the continuous data and once with the discrete data in a separate column but that gives me the following error:
Error: Discrete value supplied to continuous scale
Happy to hear suggestions!
Here's a kind of involved solution, but I believe it best answers your question, which is that you are desiring to place next to typical histogram plot a bar representing the ">10" values (or the values which are non-numeric). Critically, you want to ensure that you maintain the "binning" associated with a histogram plot, which means you are not looking to simply make your scale a discrete scale and represent a histogram with a typical barplot.
The Data
Since you want to retain histogram features, I'm going to use an example dataset that is a bit more involved than that you gave us. I'm just going to specify a uniform distribution (n=100) with 20 ">10" values thrown in there.
set.seed(123)
df<- data.frame(time=c(runif(100,0,10), rep(">10",20)))
As prepared, df$time is a character vector, but for a histogram, we need that to be numeric. We're simply going to force it to be numeric and accept that the ">10" values are going to be coerced to be NAs. This is fine, since in the end we're just going to count up those NA values and represent them with a bar. While I'm at it, I'm creating a subset of df that will be used for creating the bar representing our NAs (">10") using the count() function, which returns a dataframe consisting of one row and column: df$n = 20 in this case.
library(dplyr)
df$time <- as.numeric(df$time) #force numeric and get NA for everything else
df_na <- count(subset(df, is.na(time)))
The Plot(s)
For the actual plot, you are asking to create a combination of (1) a histogram, and (2) a barplot. These are not the same plot, but more importantly, they cannot share the same axis, since by definition, the histogram needs a continuous axis and "NA" values or ">10" is not a numeric/continuous value. The solution here is to make two separate plots, then combine them with a bit of magic thanks to cowplot.
The histogram is created quite easily. I'm saving the number of bins for demonstration purposes later. Here's the basic plot:
bin_num <- 12 # using this later
p1 <- ggplot(df, aes(x=time)) + theme_classic() +
geom_histogram(color='gray25', fill='blue', alpha=0.3, bins=bin_num)
Thanks to the subsetting previously, the barplot for the NA values is easy too:
p2 <- ggplot(df_na, aes(x=">10", y=n)) + theme_classic() +
geom_col(color='gray25', fill='red', alpha=0.3)
Yikes! That looks horrible, but have patience.
Stitching them together
You can simply run plot_grid(p1, p2) and you get something workable... but it leaves quite a lot to be desired:
There are problems here. I'll enumerate them, then show you the final code for how I address them:
Need to remove some elements from the NA barplot. Namely, the y axis entirely and the title for x axis (but it can't be NULL or the x axes won't line up properly). These are theme() elements that are easily removed via ggplot.
The NA barplot is taking up WAY too much room. Need to cut the width down. We address this by accessing the rel_widths= argument of plot_grid(). Easy peasy.
How do we know how to set the y scale upper limit? This is a bit more involved, since it will depend on the ..count.. stat for p1 as well as the numer of NA values. You can access the maximum count for a histogram using ggplot_build(), which is a part of ggplot2.
So, the final code requires the creation of the basic p1 and p2 plots, then adds to them in order to fix the limits. I'm also adding an annotation for number of bins to p1 so that we can track how well the upper limit setting works. Here's the code and some example plots where bin_num is set at 12 and 5, respectively:
# basic plots
p1 <- ggplot(df, aes(x=time)) + theme_classic() +
geom_histogram(color='gray25', fill='blue', alpha=0.3, bins=bin_num)
p2 <- ggplot(df_na, aes(x=">10", y=n)) + theme_classic() +
geom_col(color='gray25', fill='red', alpha=0.3) +
labs(x="") + theme(axis.line.y=element_blank(), axis.text.y=element_blank(),
axis.title.y=element_blank(), axis.ticks.y=element_blank()
) +
scale_x_discrete(expand=expansion(add=1))
#set upper y scale limit
max_count <- max(c(max(ggplot_build(p1)$data[[1]]$count), df_na$n))
# fix limits for plots
p1 <- p1 + scale_y_continuous(limits=c(0,max_count), expand=expansion(mult=c(0,0.15))) +
annotate('text', x=0, y=max_count, label=paste('Bins:', bin_num)) # for demo purposes
p2 <- p2 + scale_y_continuous(limits=c(0,max_count), expand=expansion(mult=c(0,0.15)))
plot_grid(p1, p2, rel_widths=c(1,0.2))
So, our upper limit fixing works. You can get really crazy playing around with positioning, etc and the plot_grid() function, but I think it works pretty well this way.
Perhaps, this is what you are looking for:
df1 <- data.frame(x=sample(1:12,50,rep=T))
df2 <- df1 %>% group_by(x) %>%
dplyr::summarise(y=n()) %>% subset(x<11)
df3 <- subset(df1, x>10) %>% dplyr::summarise(y=n()) %>% mutate(x=11)
df <- rbind(df2,df3 )
label <- ifelse((df$x<11),as.character(df$x),">10")
p <- ggplot(df, aes(x=x,y=y,color=x,fill=x)) +
geom_bar(stat="identity", position = "dodge") +
scale_x_continuous(breaks=df$x,labels=label)
p
and you get the following output:
Please note that sometimes you could have some of the bars missing depending on the sample.

How to compare two histograms in R?

I want to compare two histograms in a graph in R, but couldn't imagined and implemented.
My histograms are based on two sub-dataframes and these datasets divided according to a type (Action, Adventure Family)
My first histogram is:
split_action <- split(df, df$type)
dataset_action <- split_action$Action
hist(dataset_action$year)
split_adventure <- split(df, df$type)
dataset_adventure <- split_adventure$Adventure
hist(dataset_adventure$year)
I want to see how much overlapping is occured, their comparison based on year in the same histogram. Thank you in advence.
Using the iris dataset, suppose you want to make a histogram of sepal length for each species. First, you can make 3 data frames for each species by subsetting.
irissetosa<-subset(iris,Species=='setosa',select=c('Sepal.Length','Species'))
irisversi<-subset(iris,Species=='versicolor',select=c('Sepal.Length','Species'))
irisvirgin<-subset(iris,Species=='virginica',select=c('Sepal.Length','Species'))
and then, make the histogram for these 3 data frames. Don't forget to set the argument "add" as TRUE (for the second and third histogram), because you want to combine the histograms.
hist(irissetosa$Sepal.Length,col='red')
hist(irisversi$Sepal.Length,col='blue',add=TRUE)
hist(irisvirgin$Sepal.Length,col='green',add=TRUE)
you will have something like this
Then you can see which part is overlapping...
But, I know, it's not so good.
Another way to see which part is overlapping is by using density function.
plot(density(irissetosa$Sepal.Length),col='red')
lines(density(irisversi$Sepal.Length),col='blue')
lines(density(irisvirgin$Sepal.Length,col='green'))
Then you will have something like this
Hope it helps!!
You don't need to split the data if using ggplot. The key is to use transparency ("alpha") and change the value of the "position" argument to "identity" since the default is "stack".
Using the iris dataset:
library(ggplot2)
ggplot(data=iris, aes(x=Sepal.Length, fill=Species)) +
geom_histogram(binwidth=0.2, alpha=0.5, position="identity") +
theme_minimal()
It's not easy to see the overlap, so a density plot may be a better choice if that's the main objective. Again, use transparency to avoid obscuring overlapping plots.
ggplot(data=iris, aes(x=Sepal.Length, fill=Species)) +
geom_density(alpha=0.5) +
xlim(3.9,8.5) +
theme_minimal()
So for your data, the command would be something like this:
ggplot(data=df, aes(x=year, fill=type)) +
geom_histogram(alpha=0.5, position="identity")

Apply ggplot2 across columns

I am working with a dataframe with many columns and would like to produce certain plots of the data using ggplot2, namely, boxplots, histograms, density plots. I would like to do this by writing a single function that applies across all attributes (columns), producing one boxplot (or histogram etc) and then storing that as a given element of a list into which all the boxplots will be chained, so I could later index it by number (or by column name) in order to return the plot for a given attribute.
The issue I have is that, if I try to apply across columns with something like apply(df,2,boxPlot), I have to define boxPlot as a function that takes just a vector x. And when I do so, the attribute/column name and index are no longer retained. So e.g. in the code for producing a boxplot, like
bp <- ggplot(df, aes(x=Group, y=Attr, fill=Group)) +
geom_boxplot() +
labs(title="Plot of length per dose", x="Group", y =paste(Attr)) +
theme_classic()
the function has no idea how to extract the info necessary for Attr from just vector x (as this is just the column data and doesn't carry the column name or index).
(Note the x-axis is a factor variable called 'Group', which has 6 levels A,B,C,D,E,F, within X.)
Can anyone help with a good way of automating this procedure? (Ideally it should work for all types of ggplots; the problem here seems to simply be how to refer to the attribute name, within the ggplot function, in a way that can be applied / automatically replicated across the columns.) A for-loop would be acceptable, I guess, but if there's a more efficient/better way to do it in R then I'd prefer that!
Edit: something like what would be achieved by the top answer to this question: apply box plots to multiple variables. Except that in that answer, with his code you would still need a for-loop to change the indices on y=y[2] in the ggplot code and get all the boxplots. He's also expanded-grid to include different ````x``` possibilities (I have only one, the Group factor), but it would be easy to simplify down if the looping problem could be handled.
I'd also prefer just base R if possible--dplyr if absolutely necessary.
Here's an example of iterating over all columns of a data frame to produce a list of plots, while retaining the column name in the ggplot axis label
library(tidyverse)
plots <-
imap(select(mtcars, -cyl), ~ {
ggplot(mtcars, aes(x = cyl, y = .x)) +
geom_point() +
ylab(.y)
})
plots$mpg
You can also do this without purrr and dplyr
to_plot <- setdiff(names(mtcars), 'cyl')
plots <-
Map(function(.x, .y) {
ggplot(mtcars, aes(x = cyl, y = .x)) +
geom_point() +
ylab(.y)
}, mtcars[to_plot], to_plot)
plots$mpg

Modify Histogram to display only a fraction of total data

I hope I have not overlooked an answer to this question:
I want to make with ggplot a histogramm of only a fraction of the total data. Here's my example:
df<-iris
ggplot(data=df, aes(x=Sepal.Length, y=..density..*100)) +
geom_bar(binwidth=0.1) +
ylab("percent")
This gives a histogramm of all lines.
Now I want to limit the data passed to the plot (for instance) to a Petal.Width of 0.2. Thus the histogram I wish for, only represents the ratio "count Petal.Width=0.2 divided by total count".
Thanks for helping a ggplot-rookie!! With base plot I managed to work around, but I failed here..
I think what you want to do is to subset the data you're calling in the plot:
ggplot(data=df[df$Petal.Width == 0.2,], aes(x=Sepal.Length, y=..density..*100)) +
geom_bar(binwidth=0.1) +
ylab("percent")
Some other ways to subset data using ggplot are described in this post: Subset and ggplot2

R: Plot multiple box plots using columns from data frame

I would like to plot an INDIVIDUAL box plot for each unrelated column in a data frame. I thought I was on the right track with boxplot.matrix from the sfsmsic package, but it seems to do the same as boxplot(as.matrix(plotdata) which is to plot everything in a shared boxplot with a shared scale on the axis. I want (say) 5 individual plots.
I could do this by hand like:
par(mfrow=c(2,2))
boxplot(data$var1
boxplot(data$var2)
boxplot(data$var3)
boxplot(data$var4)
But there must be a way to use the data frame columns?
EDIT: I used iterations, see my answer.
You could use the reshape package to simplify things
data <- data.frame(v1=rnorm(100),v2=rnorm(100),v3=rnorm(100), v4=rnorm(100))
library(reshape)
meltData <- melt(data)
boxplot(data=meltData, value~variable)
or even then use ggplot2 package to make things nicer
library(ggplot2)
p <- ggplot(meltData, aes(factor(variable), value))
p + geom_boxplot() + facet_wrap(~variable, scale="free")
From ?boxplot we see that we have the option to pass multiple vectors of data as elements of a list, and we will get multiple boxplots, one for each vector in our list.
So all we need to do is convert the columns of our matrix to a list:
m <- matrix(1:25,5,5)
boxplot(x = as.list(as.data.frame(m)))
If you really want separate panels each with a single boxplot (although, frankly, I don't see why you would want to do that), I would instead turn to ggplot and faceting:
m1 <- melt(as.data.frame(m))
library(ggplot2)
ggplot(m1,aes(x = variable,y = value)) + facet_wrap(~variable) + geom_boxplot()
I used iteration to do this. I think perhaps I wasn't clear in the original question. Thanks for the responses none the less.
par(mfrow=c(2,5))
for (i in 1:length(plotdata)) {
boxplot(plotdata[,i], main=names(plotdata[i]), type="l")
}

Resources