Is there a method of converting a string iterator to R code?

Question

I am trying to calculate the mean (average) for all columns in my dataframe. I have created this code snippet

#Average overallDataset by label
overallDatasetLabels <- c("label","index","nr_pix","rows_with_2","cols_with_2","rows_with_3p","cols_with_3p","height","width","left2tile","right2tile","verticalness","top2tile","bottom2tile","horizontalness","nodiagnols")
library(dplyr)
avgOverallDataset <- summarise(group_by(overallDataset,label),nr_pix_avg=mean(nr_pix))
for (val in overallDatasetLabels){
  if (val %in% c("label","index","nr_pix")){
    next
  }
  avgOverallDataset<-cbind(avgOverallDataset,summarise(group_by(overallDataset,label),val=mean(val)))
}

When i run this code I get this error/warning:

50: In mean.default(val) : argument is not numeric or logical: returning NA

And the resulting dataframe looks like this:

The reason for this is that the val variable is being treated as a string, but I need it to be treated as "code" e.g

avgOverallDataset<- cbind(avgOverallDataset,summarise(group_by(overallDataset,label),avgrows_with_2=mean(rows_with_2)))

Would be valid.

How do I go about translating the "string" into values "in the code"?

Note: the multiple label columns can be removed using: How to remove duplicated column names in R?

I think you are looking for `mean(!!sym(val))`. See https://adv-r.hadley.nz/metaprogramming.html — Bas, Apr 13 '21 at 14:07

score 0 · Answer 1 · answered Apr 13 '21 at 14:16

Try with across so that you don't have to calculate mean for multiple columns in a loop.

cols <- setdiff(overallDatasetLabels, c("label","index","nr_pix"))

avgOverallDataset <- overallDataset %>%
                      group_by(label) %>%
                      summarise(across(all_of(cols), mean, na.rm = TRUE))

Is there a method of converting a string iterator to R code?

1 Answers1