问题
Dataframe to xts conversion fails on update from xts 0.9.7 to 0.10.0.
#THIS WORKS (uses xts 0.9.7):
library(xts)
DFX <- structure(list(DateTime = structure(list(sec = c(0, 0, 0), min = c(10L, 0L, 5L), hour = c(17L, 18L, 18L), mday = c(24L, 24L, 24L), mon = c(5L, 5L, 5L), year = c(114L, 114L, 114L), wday = c(2L, 2L, 2L), yday = c(174L, 174L, 174L), isdst = c(1L, 1L, 1L), zone = c("EDT", "EDT", "EDT"), gmtoff = c(NA_integer_, NA_integer_, NA_integer_)), .Names = c("sec", "min", "hour", "mday", "mon", "year", "wday", "yday", "isdst", "zone", "gmtoff"), class = c("POSIXlt", "POSIXt")), Open = c(125.03, 125.34, 125.85)), .Names = c("DateTime", "Open"), class = "data.frame", row.names = 558:560)
DFX <- as.xts(DFX[, -1], order.by = DFX$DateTime)
#THIS DOESN'T WORK (uses xts 0.10.1):
install.packages("devtools")
require(devtools)
install_github("joshuaulrich/xts")
library(xts)
DFX <- structure(list(DateTime = structure(list(sec = c(0, 0, 0), min = c(10L, 0L, 5L), hour = c(17L, 18L, 18L), mday = c(24L, 24L, 24L), mon = c(5L, 5L, 5L), year = c(114L, 114L, 114L), wday = c(2L, 2L, 2L), yday = c(174L, 174L, 174L), isdst = c(1L, 1L, 1L), zone = c("EDT", "EDT", "EDT"), gmtoff = c(NA_integer_, NA_integer_, NA_integer_)), .Names = c("sec", "min", "hour", "mday", "mon", "year", "wday", "yday", "isdst", "zone", "gmtoff"), class = c("POSIXlt", "POSIXt")), Open = c(125.03, 125.34, 125.85)), .Names = c("DateTime", "Open"), class = "data.frame", row.names = 558:560)
DFX <- as.xts(DFX[, -1], order.by = DFX$DateTime)
#Error message:
#Error in .Call("do_is_ordered", x = x, increasing = as.logical(increasing), :
# "do_is_ordered" not available for .Call() for package "xts"
Guess xts 0.10.0 is not a finished build. Yet I updated to 0.10.0 as quantstrat package requires blotter package, which requires xts 0.10.0. I needed this to turn a tick data dataframe into an xts object, to run a quantstrat strategy on it (requires xts objects).
回答1:
This is a problem with your R session, not xts. I'm going to assume you're running RStudio on Windows, where packages are not correctly updated if any R session is currently using the package when you attempt to update.
The solution is to close all your running R sessions, and then install xts from GitHub.
Once you do that, you will also need to change your call to as.xts, because xts objects do not like POSIXlt indexes (because they are a list of time components, not number of seconds from the epoch, like POSIXct).
R> DFX <- as.xts(DFX[, -1], order.by = DFX$DateTime)
Error in is.finite(order.by) :
default method not implemented for type 'list'
I'll open an issue to fix the error. It should be as simple as converting the POSIXlt object to POSIXct inside the as.xts call.
来源：https://stackoverflow.com/questions/44732011/xts-conversion-fails-on-update-from-xts-0-9-7-to-0-10-0