峰值以获取R中CSV的标头值

Question

I would like to set colClasses programmatically based on text in the header, but first I need to peak into the csv to get the header names that will then be used to set colClasses . How would I go about peaking? Is there a way of doing this without loading the file?

Answer 1

You can use something along the lines:

header <- read.table(csv.filename,sep=";",
                     nrows=3,header=T)

col.classes <- sapply(foo,class)
data <- as.data.frame(read.table(csv.filename,
                             sep=";",
                             header=T,
                             colClasses=col.classes))

or equivalent with read.csv

header <- read.csv(csv.filename,sep=";",
                     nrows=3,header=T)

col.classes <- sapply(foo,class)
data <- read.csv(csv.filename,
                             sep=";",
                             header=T,
                             colClasses=col.classes)