R data.table 中的双点“..”错误

Question

earlier today i was looking for a way to get a vector of the names of all columns that are class character in a data.table in R. I found this solution:

> chrs <- sapply(BTplan, is.character)
> chrs
                         plan                      plannext                     noinsnext                    claimsnext                    regionnext                        noins1 
                         TRUE                          TRUE                         FALSE                         FALSE                          TRUE                         FALSE 
                       noins2                        noins3                       region1                       region2                       region3                       claims1 
                        FALSE                         FALSE                          TRUE                          TRUE                          TRUE                         FALSE 
                      claims2                       claims3                 quotationYear               historicalYear1               historicalYear2               historicalYear3 
                        FALSE                         FALSE                         FALSE                         FALSE                         FALSE                         FALSE 
                       policy                        planNr        combined.PriceAnnually experienceBased.PriceAnnually  referenceBased.PriceAnnually                  errorMessage 
                         TRUE                          TRUE                         FALSE                         FALSE                         FALSE                          TRUE 
                         size 
                        FALSE 
> chrCols <- names(BTplan[, ..chrs])

it worked earlier today, but now, for some reason I am getting an error:

Error in `[.data.frame`(BTplan, , ..chrs) : object '..chrs' not found

What is the problem? Why does the .. no longer work?

Answer 1

Based on the error, the 'BTplan' may still be a data.frame . Converting to data.table with setDT or as.data.table should fix it as the .. and the syntax is specific for a data.table object

library(data.table)
names(as.data.table(BTplan)[, ..chrs])

---

Using a reproducible example

chrs <- sapply(iris, is.numeric)
names(as.data.table(iris)[, ..chrs])
#[1] "Sepal.Length" "Sepal.Width"  "Petal.Length" "Petal.Width"