user3148607 user3148607 - 2 months ago 36
R Question

facet_wrap for histogram in a loop: "At least one layer must contain all variables"

I am looking at producing a faceted histogram in a loop. The problem occurs when I am calling the facet_wrap within the loop. I tried different options but all of them failed with the following message:


Error in layout_base(data, vars, drop = drop) : At least one layer must contain all variables used for facetting


Below is a reproducible example.

library(ggplot2)
library(scales) #date_format

## Reproducible example

datatest <- data.frame(
column1 = sample(c("yes", "no"), 50, replace = TRUE),
column2 = sample(c("yes", "no"), 50, replace = TRUE),
column3 = sample(c("yes", "no"), 50, replace = TRUE),
column4 = sample(c("yes", "no"), 50, replace = TRUE)
)

#This function will generate a uniform sample of dates from
#within a designated start and end date:
rand.date=function(start.day,end.day,data){
size=dim(data)[1]
days=seq.Date(as.Date(start.day),as.Date(end.day),by="day")
pick.day=runif(size,1,length(days))
date=days[pick.day]
}

#This will create a new column within your data frame called date:
datatest$date=rand.date("2016-01-01","2016-09-21",datatest)

## Simple frequency plot that works well
histotest <- ggplot(datatest, aes(x = date)) +
geom_histogram(binwidth = 7, fill="#2a87c8", colour="white") +
scale_x_date(limits = c(Sys.Date() - 250, NA), labels = date_format("%b %Y")) +
labs(x = "Period", y = "Count") +
facet_wrap(~ column1 , ncol=1) +
theme(plot.title=element_text(face="bold", size=9),
panel.grid.major = element_line(colour = "white"),
panel.grid.minor = element_blank())
ggsave("out/column1_histo.png", plot=histotest, width=12, height=6,units="in", dpi=300)


Here comes the problem:

## Same plot generated through a loop
## The pb is with the facet_wrap
for (i in 1:4 ) {
rm(variablename)
variablename <- names(datatest)[i]
## histogramme to display event occurence over time
histoloop <- ggplot(datatest, aes(x = date)) +
geom_histogram(binwidth = 7, fill="#2a87c8", colour="white") +
scale_x_date(limits = c(Sys.Date() - 250, NA), labels = date_format("%b %Y")) +
labs(x = "Period", y = "Count") +

## I tried different options but none of them is working
## If I comment the facet_wrap everything's fine...
facet_wrap(~ variablename , ncol=1) +
#facet_wrap(~ names(datatest)[i] , ncol=1) +
#facet_wrap(~ aes_string(names(datatest)[i]) , ncol=1) +
theme(plot.title=element_text(face="bold", size=9),
panel.grid.major = element_line(colour = "white"),
panel.grid.minor = element_blank())
ggsave(filename=paste("out/",variablename,"_histo.png",sep=""), plot=histoloop, width=12, height=6,units="in", dpi=300)
}

Answer

Reformat data for plotting, then plot with split data in a loop:

library(tidyr) #gather

# prepare the data, wide to long
plotDat <- gather(datatest, key = "Column", value = "Value", -c(date) )

# then use loop, I prefer lapply
lapply(split(plotDat, plotDat$Column), function(i){
  ggplot(i, aes(x = date)) + 
    geom_histogram(binwidth = 7, fill="#2a87c8", colour="white") +
    scale_x_date(limits = c(Sys.Date() - 250, NA), labels = date_format("%b %Y")) +
    labs(x = "Period", y = "Count") +
    facet_wrap(~ Value , ncol = 1) +
    theme(plot.title=element_text(face="bold", size=9),
          panel.grid.major = element_line(colour = "white"),
          panel.grid.minor = element_blank())
})