Reliable way to detect if a column in a data.frame is.POSIXct

问题

R has is.vector, is.list, is.integer, is.double, is.numeric, is.factor, is.character, etc. Why is there no is.POSIXct, is.POSIXlt or is.Date?

I need a reliable way to detect POSIXct object, and class(x)[1] == "POSIXct" seems really... dirty.

回答1:

I would personally just use inherits as joran suggested. You could use it to create your own is.POSIXct function.

# functions
is.POSIXct <- function(x) inherits(x, "POSIXct")
is.POSIXlt <- function(x) inherits(x, "POSIXlt")
is.POSIXt <- function(x) inherits(x, "POSIXt")
is.Date <- function(x) inherits(x, "Date")
# data
d <- data.frame(pct = Sys.time())
d$plt <- as.POSIXlt(d$pct)
d$date <- Sys.Date()
# checks
sapply(d, is.POSIXct)
#   pct   plt  date 
#  TRUE FALSE FALSE 
sapply(d, is.POSIXlt)
#   pct   plt  date 
# FALSE  TRUE FALSE 
sapply(d, is.POSIXt)
#   pct   plt  date 
#  TRUE  TRUE FALSE 
sapply(d, is.Date)
#   pct   plt  date 
# FALSE FALSE  TRUE 

回答2:

The lubridate package has is.POSIXt, is.POSIXct, is.POSIXlt, and is.Date functions.

回答3:

You can try is(). This is what the lubridate functions is.Date and is.POSIX* rely on anyway.

x <- Sys.time()
class(x)
# [1] "POSIXct" "POSIXt" 
is(x, "Date")
#v[1] FALSE
is(x, "POSIXct")
# [1] TRUE

y <- Sys.Date()
class(y)
# [1] "Date"
is(y, "POSIXct")
# [1] FALSE
is(y, "Date")
# [1] TRUE

来源：https://stackoverflow.com/questions/26412947/reliable-way-to-detect-if-a-column-in-a-data-frame-is-posixct