Debug a function in R that returns NA

Question

I am trying to write a function where it reads some pollution data from different files in a given directory, each file corresponding to records of a monitor and returns the mean value for one of the pollutants given as input argument to the function.

The data looks like this:

pollutantmean <- function(directory, pollutant, ID){
        pmean <- 0
        for (monitor in ID){
                filename <- paste("./W2/", directory,"/", monitor,'.csv', sep = "")
                print(filename)
                pollution <- read.csv(file = filename, header = TRUE, sep = ",")
                class(pollution)
                head(pollution, n = 3)
                pmean <- pmean + mean(pollution$pollutant, na.rm = TRUE)
        }
        return(pmean)
}

To run the function I set:

pollutantmean("specdata", "sulfate", 110:112)

I am getting the following output and error message:

[1] "./W2/specdata/110.csv"
[1] NA
Warning message:
In mean.default(pollution$pollutant, na.rm = TRUE) :
  argument is not numeric or logical: returning NA 

My questions are:

1. How can I run the code line by line such that I see how the value of the local variables in the function changes? I looked up Youtube, but couldn't find a relevant video about it.
2. When I copy-paste and run each line of the code in the function in console I see that the data is stored properly in the pollution data.frame. why can't I see the class and the head of the data.frame printed as output in the console? Instead I only get NA which is related to pmean I guess...
3. How can I fix the error related to this error message?

question from:https://stackoverflow.com/questions/66051706/debug-a-function-in-r-that-returns-na

Answer 1

Good question. If you are using RStudio you can use the browser() command to freeze the program at a certain spot so you can view the environment.

Be warned that it will stop every time you hit browser() so you might need to put an if statement to detect a warning before running it. And don't forget to remove the statement once you're done.

pollutantmean <- function(directory, pollutant, ID){
  pmean <- 0
  for (monitor in ID){
    filename <- paste("./W2/", directory,"/", monitor,'.csv', sep = "")
    print(filename)
    pollution <- read.csv(file = filename, header = TRUE, sep = ",")
    class(pollution)
    head(pollution, n = 3)
    pmean <- pmean + mean(pollution$pollutant, na.rm = TRUE)
    # Stop here and view the current environment
    browser()
  }
  return(pmean)
}