Related
This question already has answers here:
Plotting two variables as lines using ggplot2 on the same graph
(5 answers)
Closed 8 months ago.
I am new to R and have the following example code that I wish to apply for every column in my data.
data(economics, package="ggplot2")
economics$index <- 1:nrow(economics)
loessMod10 <- loess(uempmed ~ index, data=economics, span=0.10)
smoothed10 <- predict(loessMod10)
plot(economics$uempmed, x=economics$date, type="l", main="Loess Smoothing and Prediction", xlab="Date", ylab="Unemployment (Median)")
lines(smoothed10, x=economics$date, col="red")
Could someone please suggest how this would be possible?
It's possible to perform loess smoothing within ggplot.
library(data.table)
library(ggplot2)
df <- economics
##
#
gg.melt <- setDT(df) |> melt(id='date', variable.name = 'KPI')
ggplot(gg.melt, aes(x=date, y=value))+
geom_line()+
stat_smooth(method=loess, color='red', size=0.5, se=FALSE, method.args = list(span=0.1))+
facet_wrap(~KPI, scales = 'free_y')
Regarding combining everything on one plot I'm not seeing how you would do that as the y-scales are so different. If the point is to see how the peaks line up, etc. you could do this:
ggplot(gg.melt, aes(x=date, y=value))+
geom_line()+
stat_smooth(method=loess, color='red', size=0.5, se=FALSE, method.args = list(span=0.1))+
facet_grid(KPI~., scales = 'free_y')
There is also the dygraphs package which allows creation of dynamic graphics that can be saved to html:
gg.melt[, scaled:=scale(value, center = FALSE, scale=diff(range(value))), by=.(KPI)]
gg.melt[, pred:=predict(loess(scaled~as.integer(date), .SD, span=0.1)), by=.(KPI)]
gg.dt <- dcast(gg.melt, date~KPI, value.var = list('scaled', 'pred'))
library(dygraphs)
dygraph(gg.dt) |>
dyCrosshair(direction = 'vertical') |>
dyRangeSelector()
It's possible to create a dygraph(...) version of the second plot, where the different KPI are in different facets, but you have to use RMarkdown for that.
You can make your data from wide to long by the date and use facet_wrap. Maybe you want something like this:
library(ggplot2)
library(reshape2)
library(dplyr)
economics %>%
melt(., "date") %>%
ggplot(., aes(date, value)) +
geom_line() +
facet_wrap(~variable, scales = "free")
Output:
Comment: All plots in one graph
If you mean all plots in one graph, you can give the variables a color like this:
economics %>%
melt(., "date") %>%
ggplot(., aes(date, value, color = variable)) +
geom_line() +
scale_y_log10()
Output:
I’m totally new to ggplot, relatively fresh with R and want to make a smashing ”before-and-after” scatterplot with connecting lines to illustrate the movement in percentages of different subgroups before and after a special training initiative. I’ve tried some options, but have yet to:
show each individual observation separately (now same values are overlapping)
connect the related before and after measures (x=0 and X=1) with lines to more clearly illustrate the direction of variation
subset the data along class and id using shape and colors
How can I best create a scatter plot using ggplot (or other) fulfilling the above demands?
Main alternative: geom_point()
Here is some sample data and example code using genom_point
x <- c(0,0,0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1,1,1) # 0=before, 1=after
y <- c(45,30,10,40,10,NA,30,80,80,NA,95,NA,90,NA,90,70,10,80,98,95) # percentage of ”feelings of peace"
class <- c(0,0,0,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,1,1) # 0=multiple days 1=one day
id <- c(1,1,2,3,4,4,4,4,5,6,1,1,2,3,4,4,4,4,5,6) # id = per individual
df <- data.frame(x,y,class,id)
ggplot(df, aes(x=x, y=y), fill=id, shape=class) + geom_point()
Alternative: scale_size()
I have explored stat_sum() to summarize the frequencies of overlapping observations, but then not being able to subset using colors and shapes due to overlap.
ggplot(df, aes(x=x, y=y)) +
stat_sum()
Alternative: geom_dotplot()
I have also explored geom_dotplot() to clarify the overlapping observations that arise from using genom_point() as I do in the example below, however I have yet to understand how to combine the before and after measures into the same plot.
df1 <- df[1:10,] # data before
df2 <- df[11:20,] # data after
p1 <- ggplot(df1, aes(x=x, y=y)) +
geom_dotplot(binaxis = "y", stackdir = "center",stackratio=2,
binwidth=(1/0.3))
p2 <- ggplot(df2, aes(x=x, y=y)) +
geom_dotplot(binaxis = "y", stackdir = "center",stackratio=2,
binwidth=(1/0.3))
grid.arrange(p1,p2, nrow=1) # GridExtra package
Or maybe it is better to summarize data by x, id, class as mean/median of y, filter out ids producing NAs (e.g. ids 3 and 6), and connect the points by lines? So in case if you don't really need to show variability for some ids (which could be true if the plot only illustrates tendencies) you can do it this way:
library(ggplot)
library(dplyr)
#library(ggthemes)
df <- df %>%
group_by(x, id, class) %>%
summarize(y = median(y, na.rm = T)) %>%
ungroup() %>%
mutate(
id = factor(id),
x = factor(x, labels = c("before", "after")),
class = factor(class, labels = c("one day", "multiple days")),
) %>%
group_by(id) %>%
mutate(nas = any(is.na(y))) %>%
ungroup() %>%
filter(!nas) %>%
select(-nas)
ggplot(df, aes(x = x, y = y, col = id, group = id)) +
geom_point(aes(shape = class)) +
geom_line(show.legend = F) +
#theme_few() +
#theme(legend.position = "none") +
ylab("Feelings of peace, %") +
xlab("")
Here's one possible solution for you.
First - to get the color and shapes determined by variables, you need to put these into the aes function. I turned several into factors, so the labs function fixes the labels so they don't appear as "factor(x)" but just "x".
To address multiple points, one solution is to use geom_smooth with method = "lm". This plots the regression line, instead of connecting all the dots.
The option se = FALSE prevents confidence intervals from being plotted - I don't think they add a lot to your plot, but play with it.
Connecting the dots is done by geom_line - feel free to try that as well.
Within geom_point, the option position = position_jitter(width = .1) adds random noise to the x-axis so points do not overlap.
ggplot(df, aes(x=factor(x), y=y, color=factor(id), shape=factor(class), group = id)) +
geom_point(position = position_jitter(width = .1)) +
geom_smooth(method = 'lm', se = FALSE) +
labs(
x = "x",
color = "ID",
shape = 'Class'
)
I am making a dodged barplot in ggplot2 and one grouping has a zero count that I want to display. I remembered seeing this on HERE a while back and figured the scale_x_discrete(drop=F) would work. It does not appear to work with dodged bars. How can I make the zero counts show?
For instance, (code below) in the plot below, type8~group4 has no examples. I would still like the plot to display the empty space for the zero count instead of eliminating the bar. How can I do this?
mtcars2 <- data.frame(type=factor(mtcars$cyl),
group=factor(mtcars$gear))
m2 <- ggplot(mtcars2, aes(x=type , fill=group))
p2 <- m2 + geom_bar(colour="black", position="dodge") +
scale_x_discrete(drop=F)
p2
Here's how you can do it without making summary tables first.
It did not work in my CRAN versioin (2.2.1) but in the latest development version of ggplot (2.2.1.900) I had no issues.
ggplot(mtcars, aes(factor(cyl), fill = factor(vs))) +
geom_bar(position = position_dodge(preserve = "single"))
http://ggplot2.tidyverse.org/reference/position_dodge.html
Updated geom_bar() needs stat = "identity"
For what it's worth: The table of counts, dat, above contains NA. Sometimes, it is useful to have an explicit 0 instead; for instance, if the next step is to put counts above the bars. The following code does just that, although it's probably no simpler than Joran's. It involves two steps: get a crosstabulation of counts using dcast, then melt the table using melt, followed by ggplot() as usual.
library(ggplot2)
library(reshape2)
mtcars2 = data.frame(type=factor(mtcars$cyl), group=factor(mtcars$gear))
dat = dcast(mtcars2, type ~ group, fun.aggregate = length)
dat.melt = melt(dat, id.vars = "type", measure.vars = c("3", "4", "5"))
dat.melt
ggplot(dat.melt, aes(x = type,y = value, fill = variable)) +
geom_bar(stat = "identity", colour = "black", position = position_dodge(width = .8), width = 0.7) +
ylim(0, 14) +
geom_text(aes(label = value), position = position_dodge(width = .8), vjust = -0.5)
The only way I know of is to pre-compute the counts and add a dummy row:
dat <- rbind(ddply(mtcars2,.(type,group),summarise,count = length(group)),c(8,4,NA))
ggplot(dat,aes(x = type,y = count,fill = group)) +
geom_bar(colour = "black",position = "dodge",stat = "identity")
I thought that using stat_bin(drop = FALSE,geom = "bar",...) instead would work, but apparently it does not.
I asked this same question, but I only wanted to use data.table, as it's a faster solution for much larger data sets. I included notes on the data so that those that are less experienced and want to understand why I did what I did can do so easily. Here is how I manipulated the mtcars data set:
library(data.table)
library(scales)
library(ggplot2)
mtcars <- data.table(mtcars)
mtcars$Cylinders <- as.factor(mtcars$cyl) # Creates new column with data from cyl called Cylinders as a factor. This allows ggplot2 to automatically use the name "Cylinders" and recognize that it's a factor
mtcars$Gears <- as.factor(mtcars$gear) # Just like above, but with gears to Gears
setkey(mtcars, Cylinders, Gears) # Set key for 2 different columns
mtcars <- mtcars[CJ(unique(Cylinders), unique(Gears)), .N, allow.cartesian = TRUE] # Uses CJ to create a completed list of all unique combinations of Cylinders and Gears. Then counts how many of each combination there are and reports it in a column called "N"
And here is the call that produced the graph
ggplot(mtcars, aes(x=Cylinders, y = N, fill = Gears)) +
geom_bar(position="dodge", stat="identity") +
ylab("Count") + theme(legend.position="top") +
scale_x_discrete(drop = FALSE)
And it produces this graph:
Furthermore, if there is continuous data, like that in the diamonds data set (thanks to mnel):
library(data.table)
library(scales)
library(ggplot2)
diamonds <- data.table(diamonds) # I modified the diamonds data set in order to create gaps for illustrative purposes
setkey(diamonds, color, cut)
diamonds[J("E",c("Fair","Good")), carat := 0]
diamonds[J("G",c("Premium","Good","Fair")), carat := 0]
diamonds[J("J",c("Very Good","Fair")), carat := 0]
diamonds <- diamonds[carat != 0]
Then using CJ would work as well.
data <- data.table(diamonds)[,list(mean_carat = mean(carat)), keyby = c('cut', 'color')] # This step defines our data set as the combinations of cut and color that exist and their means. However, the problem with this is that it doesn't have all combinations possible
data <- data[CJ(unique(cut),unique(color))] # This functions exactly the same way as it did in the discrete example. It creates a complete list of all possible unique combinations of cut and color
ggplot(data, aes(color, mean_carat, fill=cut)) +
geom_bar(stat = "identity", position = "dodge") +
ylab("Mean Carat") + xlab("Color")
Giving us this graph:
Use count and complete from dplyr to do this.
library(tidyverse)
mtcars %>%
mutate(
type = as.factor(cyl),
group = as.factor(gear)
) %>%
count(type, group) %>%
complete(type, group, fill = list(n = 0)) %>%
ggplot(aes(x = type, y = n, fill = group)) +
geom_bar(colour = "black", position = "dodge", stat = "identity")
You can exploit the feature of the table() function, which computes the number of occurrences of a factor for all its levels
# load plyr package to use ddply
library(plyr)
# compute the counts using ddply, including zero occurrences for some factor levels
df <- ddply(mtcars2, .(group), summarise,
types = as.numeric(names(table(type))),
counts = as.numeric(table(type)))
# plot the results
ggplot(df, aes(x = types, y = counts, fill = group)) +
geom_bar(stat='identity',colour="black", position="dodge")
I am making a dodged barplot in ggplot2 and one grouping has a zero count that I want to display. I remembered seeing this on HERE a while back and figured the scale_x_discrete(drop=F) would work. It does not appear to work with dodged bars. How can I make the zero counts show?
For instance, (code below) in the plot below, type8~group4 has no examples. I would still like the plot to display the empty space for the zero count instead of eliminating the bar. How can I do this?
mtcars2 <- data.frame(type=factor(mtcars$cyl),
group=factor(mtcars$gear))
m2 <- ggplot(mtcars2, aes(x=type , fill=group))
p2 <- m2 + geom_bar(colour="black", position="dodge") +
scale_x_discrete(drop=F)
p2
Here's how you can do it without making summary tables first.
It did not work in my CRAN versioin (2.2.1) but in the latest development version of ggplot (2.2.1.900) I had no issues.
ggplot(mtcars, aes(factor(cyl), fill = factor(vs))) +
geom_bar(position = position_dodge(preserve = "single"))
http://ggplot2.tidyverse.org/reference/position_dodge.html
Updated geom_bar() needs stat = "identity"
For what it's worth: The table of counts, dat, above contains NA. Sometimes, it is useful to have an explicit 0 instead; for instance, if the next step is to put counts above the bars. The following code does just that, although it's probably no simpler than Joran's. It involves two steps: get a crosstabulation of counts using dcast, then melt the table using melt, followed by ggplot() as usual.
library(ggplot2)
library(reshape2)
mtcars2 = data.frame(type=factor(mtcars$cyl), group=factor(mtcars$gear))
dat = dcast(mtcars2, type ~ group, fun.aggregate = length)
dat.melt = melt(dat, id.vars = "type", measure.vars = c("3", "4", "5"))
dat.melt
ggplot(dat.melt, aes(x = type,y = value, fill = variable)) +
geom_bar(stat = "identity", colour = "black", position = position_dodge(width = .8), width = 0.7) +
ylim(0, 14) +
geom_text(aes(label = value), position = position_dodge(width = .8), vjust = -0.5)
The only way I know of is to pre-compute the counts and add a dummy row:
dat <- rbind(ddply(mtcars2,.(type,group),summarise,count = length(group)),c(8,4,NA))
ggplot(dat,aes(x = type,y = count,fill = group)) +
geom_bar(colour = "black",position = "dodge",stat = "identity")
I thought that using stat_bin(drop = FALSE,geom = "bar",...) instead would work, but apparently it does not.
I asked this same question, but I only wanted to use data.table, as it's a faster solution for much larger data sets. I included notes on the data so that those that are less experienced and want to understand why I did what I did can do so easily. Here is how I manipulated the mtcars data set:
library(data.table)
library(scales)
library(ggplot2)
mtcars <- data.table(mtcars)
mtcars$Cylinders <- as.factor(mtcars$cyl) # Creates new column with data from cyl called Cylinders as a factor. This allows ggplot2 to automatically use the name "Cylinders" and recognize that it's a factor
mtcars$Gears <- as.factor(mtcars$gear) # Just like above, but with gears to Gears
setkey(mtcars, Cylinders, Gears) # Set key for 2 different columns
mtcars <- mtcars[CJ(unique(Cylinders), unique(Gears)), .N, allow.cartesian = TRUE] # Uses CJ to create a completed list of all unique combinations of Cylinders and Gears. Then counts how many of each combination there are and reports it in a column called "N"
And here is the call that produced the graph
ggplot(mtcars, aes(x=Cylinders, y = N, fill = Gears)) +
geom_bar(position="dodge", stat="identity") +
ylab("Count") + theme(legend.position="top") +
scale_x_discrete(drop = FALSE)
And it produces this graph:
Furthermore, if there is continuous data, like that in the diamonds data set (thanks to mnel):
library(data.table)
library(scales)
library(ggplot2)
diamonds <- data.table(diamonds) # I modified the diamonds data set in order to create gaps for illustrative purposes
setkey(diamonds, color, cut)
diamonds[J("E",c("Fair","Good")), carat := 0]
diamonds[J("G",c("Premium","Good","Fair")), carat := 0]
diamonds[J("J",c("Very Good","Fair")), carat := 0]
diamonds <- diamonds[carat != 0]
Then using CJ would work as well.
data <- data.table(diamonds)[,list(mean_carat = mean(carat)), keyby = c('cut', 'color')] # This step defines our data set as the combinations of cut and color that exist and their means. However, the problem with this is that it doesn't have all combinations possible
data <- data[CJ(unique(cut),unique(color))] # This functions exactly the same way as it did in the discrete example. It creates a complete list of all possible unique combinations of cut and color
ggplot(data, aes(color, mean_carat, fill=cut)) +
geom_bar(stat = "identity", position = "dodge") +
ylab("Mean Carat") + xlab("Color")
Giving us this graph:
Use count and complete from dplyr to do this.
library(tidyverse)
mtcars %>%
mutate(
type = as.factor(cyl),
group = as.factor(gear)
) %>%
count(type, group) %>%
complete(type, group, fill = list(n = 0)) %>%
ggplot(aes(x = type, y = n, fill = group)) +
geom_bar(colour = "black", position = "dodge", stat = "identity")
You can exploit the feature of the table() function, which computes the number of occurrences of a factor for all its levels
# load plyr package to use ddply
library(plyr)
# compute the counts using ddply, including zero occurrences for some factor levels
df <- ddply(mtcars2, .(group), summarise,
types = as.numeric(names(table(type))),
counts = as.numeric(table(type)))
# plot the results
ggplot(df, aes(x = types, y = counts, fill = group)) +
geom_bar(stat='identity',colour="black", position="dodge")
I'm looking for a way to plot a bar chart containing two different series, hide the bars for one of the series and instead have a line (smooth if possible) go through the top of where bars for the hidden series would have been (similar to how one might overlay a freq polynomial on a histogram). I've tried the example below but appear to be running into two problems.
First, I need to summarize (total) the data by group, and second, I'd like to convert one of the series (df2) to a line.
df <- data.frame(grp=c("A","A","B","B","C","C"),val=c(1,1,2,2,3,3))
df2 <- data.frame(grp=c("A","A","B","B","C","C"),val=c(1,4,3,5,1,2))
ggplot(df, aes(x=grp, y=val)) +
geom_bar(stat="identity", alpha=0.75) +
geom_bar(data=df2, aes(x=grp, y=val), stat="identity", position="dodge")
You can get group totals in many ways. One of them is
with(df, tapply(val, grp, sum))
For simplicity, you can combine bar and line data into a single dataset.
df_all <- data.frame(grp = factor(levels(df$grp)))
df_all$bar_heights <- with(df, tapply(val, grp, sum))
df_all$line_y <- with(df2, tapply(val, grp, sum))
Bar charts use a categorical x-axis. To overlay a line you will need to convert the axis to be numeric.
ggplot(df_all) +
geom_bar(aes(x = grp, weight = bar_heights)) +
geom_line(aes(x = as.numeric(grp), y = line_y))
Perhaps your sample data aren't representative of the real data you are working with, but there are no lines to be drawn for df2. There is only one value for each x and y value. Here's a modifed version of your df2 with enough data points to construct lines:
df <- data.frame(grp=c("A","A","B","B","C","C"),val=c(1,2,3,1,2,3))
df2 <- data.frame(grp=c("A","A","B","B","C","C"),val=c(1,4,3,5,0,2))
p <- ggplot(df, aes(x=grp, y=val))
p <- p + geom_bar(stat="identity", alpha=0.75)
p + geom_line(data=df2, aes(x=grp, y=val), colour="blue")
Alternatively, if your example data above is correct, you can plot this information as a point with geom_point(data = df2, aes(x = grp, y = val), colour = "red", size = 6). You can obviously change the color and size to your liking.
EDIT: In response to comment
I'm not entirely sure what the visual for a freq polynomial over a histogram is supposed to look like. Are the x-values supposed to be connected to one another? Secondly, you keep referring to wanting lines but your code shows geom_bar() which I assume isn't what you want? If you want lines, use geom_lines(). If the two assumptions above are correct, then here's an approach to do that:
#First let's summarise df2 by group
df3 <- ddply(df2, .(grp), summarise, total = sum(val))
> df3
grp total
1 A 5
2 B 8
3 C 3
#Second, let's plot df3 as a line while treating the grp variable as numeric
p <- ggplot(df, aes(x=grp, y=val))
p <- p + geom_bar(alpha=0.75, stat = "identity")
p + geom_line(data=df3, aes(x=as.numeric(grp), y=total), colour = "red")