Converting elements in multiple variables by for loop is not working

I have ran code to save names of variables which are character and binary in object binvar

print(binvar)

[1] "x1" "x2" "x3" "x4"
[5] "x5"

I want to use this to select columns in my the original dataframe df in order to run a for loop where I convert yes and no to 1 and 0.

for(i in 1:length(binvar)){
  for(j in 1:length(nrow(df))){
    if(df[[binvar[i]]][j]=="Yes"){
      df[[binvar[i]]][j]<-1
    }
    else if(df[[binvar[i]]][j]=="No"){
      df[[binvar[i]]][j]<-0
    }
  }
}

The problem is that the loop doesn’t iterate over all elements inside the selected columns but only on the first element of each column. I thought iterator j would go over all element.

How can I solve this?

>Solution :

We don’t need a nested loop in R. The == is elementwise comparison operator. Subset the dataset (‘df’) based on the vector of column names (‘binvar’) i.e. df[binvar]), create a logical matrix (== "Yes") and coerce the logical to binary (+) and assign (<-) back to the subset of columns (assuming there are only ‘Yes’, ‘No’ and possibly NA)

df[binvar] <- +(df[binvar] == "Yes")

In the OP’s nested loop, the second expression is the problematic because nrow returns a single value and the length will be 1, so 1:1 is always 1.

for(j in 1:length(nrow(df)))

It should be

for(j in seq_len(nrow(df)))

If we want a reproducible example

> data(iris)
> nrow(iris)
[1] 150
> length(nrow(iris))
[1] 1
> 1:length(nrow(iris))
[1] 1