Clever Geek Handbook

Argument is not numeric or boolean: return NA to R

I am trying to write a function to calculate the average of a column. The function has an argument directory and a column_name. However, I keep getting the error "argument is not numeric or boolean: return NA":

pollutantmean <- function(directory, pollutant) {

    directoryVal <- directory
    pollutantVal <- pollutant

    pollutantData <- read.csv(directoryVal)
    meanVal <- mean(pollutantData$pollutantVal, na.rm = TRUE)

}

I named it:

pollutantmean("001.csv", "nitrate")

"nitrate" is one of the column names.

Note that the following works, so I'm not sure why it doesn't work in my function:

mydata <- read.csv("001.csv")
mean(mydata$nitrate, na.rm = TRUE)

Please, help. Thank.

+3


source to share


2 answers


You should try meanVal <- mean(pollutantData[pollutantVal], na.rm = TRUE)

, in reality your dataframe pollutantData

doesn't have any call to the column pollutantVall

, so you have pollutantData$pollutantVal

which is NULL

. If you want to access a column of a data frame using a character, you must use square brackets.



+2


source


I want to add a fix. In fact it should be

meanVal <- mean(pollutantData[,pollutantVal], na.rm = TRUE)


Notice ,

before pollutantVal

, as it pollutantVal

is a column and must be indexed as such.

+3


source






More articles:


All Articles