R attributes() attr()

兩個函數實際上就是為對象創建屬性

attributes()

讀取對象屬性

x<- table(c(3,4,6,8,6))
attributes(x)
$dim
[1] 4

$dimnames
$dimnames[[1]]
[1] "3" "4" "6" "8"


$class
[1] "table"

attributes()函數修改屬性

如

attributes(x) <- NULL
x
## [1] 1 1 2 1

如上修改后x不再是數組，也不是table

attr函數

可以用attr(x, "屬性名")的格式讀取或定義x的屬性。 如：

x <- c(1,3,5)
attr(x, "theta") <- c(0, 1)
print(x)
## [1] 1 3 5
## attr(,"theta")
## [1] 0 1

可以讓向量x額外地保存一個theta屬性， 這樣的屬性常常成為“元數據”(meta data)， 比如， 用來保存數據的說明、模擬數據的真實模型參數，等等

對比

x<-c(apple=2.5,orange=2.1)

attributes(x)
$names
[1] "apple" "orange"

attr(x,"names")
[1] "apple" "orange"
attr(x,"names")<-c("apple","grape")
attr(x,"type")<-"fruit"
x
apple grape
2.5 2.1
attr(,"type")
[1] "fruit"
attributes(x)
$names
[1] "apple" "grape"

$type
[1] "fruit"