Create a disk.frame containing only specified columns

[lucazav]

Is there a simply way to create a disk.frame containing only specified columns?
I suppose it could be done using the csv_to_disk.frame's inmapfn parameter in the following way:

inmapfn = function(chunk) {
cols_to_keep = c('col01', 'col02', ...)
chunk[, (names(chunk) %in% cols_to_keep)]
...
}

or something like that,
Maybe a specific parameter like colsToKeep into the cst_to_disk.frame function could be really useful.

[xiaodaigh]

You can already do that since {disk.frame} uses {data.table}'s fread which provides the select= option

library(disk.frame)
#setup_disk.frame() # omitted since the data is tiny

# write the iris dataset to csv
data.table::fwrite(iris, "iris.csv")

# read only the Species and Sepal.Length columns via `select = `
iris.df = disk.frame::csv_to_disk.frame("iris.csv", select = c("Species", "Sepal.Length"))


collect(iris.df)

# Species Sepal.Length
# 1:    setosa          5.1
# 2:    setosa          4.9
# 3:    setosa          4.7
# 4:    setosa          4.6
# 5:    setosa          5.0
# ---                       
#   146: virginica          6.7
# 147: virginica          6.3
# 148: virginica          6.5
# 149: virginica          6.2
# 150: virginica          5.9