R Passing arguments for ggplot through a function call with facet_wrap and data subsetting

Question

I have a longitudinal data which I pass to ggplot with facet_wrap and a subsetting of the dataframe. I wish to 'function'-alize this and I am running into trouble. I have seen similar posts to this, but not one with facet_wrap and data subsetting inside the function. For example, I have used information from this post in the past to do simple graphs. Below I show the section of code for generating dummy data and then plotting the graphic. That works OK. It's when I try to use the function call that I get the error message:

Your help is greatly appreciated. Thank you!!

Error: Faceting variables must have at least one value

# Test

# Generate data
DF <- expand.grid(Time=(0:10), variable=as.factor(c("ux0", "ux1", "ux2")), model=as.factor(c("Model 1", "Model 2", "Model 3")))
DF$value <- DF$Time + rnorm(nrow(DF), 0, 20)

# load libraries
library(ggplot2)
library(scales)

# Define themes
My_Theme = theme_bw()+
  theme(panel.grid.major = element_blank(), 
        panel.grid.minor = element_blank(),
        aspect.ratio = 1,
        axis.title=element_text(size=7),
        axis.text.x=element_text(size=rel(0.6)),
        axis.text.y=element_text(size=rel(0.6)),
        strip.text = element_text(size = 6))

#Plot
my.plot  =  
  ggplot(subset(DF, variable %in% "ux1")) +
  geom_line(aes(x=Time, y=value)) +
  facet_wrap( ~ model, ncol=3) + 
  labs(x = "Time [s]", y = expression(paste("U"[X],","[1]))) +
  My_Theme
print(my.plot)

#Now try with function
makePlots <- function(data, subSetVar, subSetval, xVar, yVar, facetVar, 
                      xLabel, yLabel){

  # Common Theme to all plots

  My_Theme = theme_bw()+
    theme(panel.grid.major = element_blank(), 
          panel.grid.minor = element_blank(),
          aspect.ratio = 1,
          axis.title=element_text(size=7),
          axis.text.x=element_text(size=rel(0.6)),
          axis.text.y=element_text(size=rel(0.6)),
          strip.text = element_text(size = 6))

  my.plot  =  
    ggplot(subset(data, subSetVar %in% subSetval)) + 
    geom_line(aes(x=xVar, y=yVar)) + 
    facet_wrap(facetVar, ncol=3) + 
    labs(x = xLabel, y = yLabel) +
    My_Theme


  # Output to Plots window in RStudio
  print(my.plot)

}

my.plot <- makePlots(DF, "variable", "ux1", "Time", "value", "model",
                     "Time [s]", expression(paste("U"[X],","[1])))

'''

dc37 · Accepted Answer · 2020-04-01T23:49:02.327

In order to pass character strings as variable into ggplot, you need to make some changes in my.plot part before wrapping it into a function.

For subset the dataset, you need to pass the names of the column with [[ ]] in order to make it work. For the definition of x and y, you can use aes_string (https://ggplot2.tidyverse.org/reference/aes_.html). Finally, for facetting, pass your character vector as a formula (as explained in this post: Passing string variable facet_wrap() in ggplot using R).

my.plot  =  
    ggplot(subset(data, data[[subSetVar]] %in% subSetval)) + 
    geom_line(aes_string(x=xVar, y=yVar)) +
    facet_wrap(as.formula(paste("~", facetVar)), ncol=3) + 
    labs(x = xLabel, y = yLabel) +
    My_Theme

Then, it should work and get you the same graph :

my.plot <- makePlots(DF, "variable", "ux1", "Time", "value", "model",
                     "Time [s]", expression(paste("U"[X],","[1])))

Does it answer your question ?

Thank you very much indeed. That worked very nicely! – please help Apr 02 '20 at 00:32 — please help, Apr 02 '20 at 00:32

R Passing arguments for ggplot through a function call with facet_wrap and data subsetting

1 Answers1