Skip to content

Class for time series

5 messages · Poizot Emmanuel, Hasan Diwan, Achim Zeileis +2 more

#
Dear all,

I have a time serie dataset such as the following with data acquired 
every 15 minutes:

Date    Heure Profondeur Temp?rature Salinit? Turbidit? Chloration
1 2012-07-06 08:47:22     -0.144      22.469    0.011 0.000          0
2 2012-07-06 09:02:21     -0.147      22.476    0.011 0.000          0
3 2012-07-06 09:17:21     -0.139      22.498    0.011 19.323          0
4 2012-07-06 09:32:21     -0.136      22.540    0.011 19.343          0
5 2012-07-06 09:47:21     -0.141      22.510    0.011 19.321          0
6 2012-07-06 10:02:21     -0.139      22.372    0.011 19.280          0

I wonder what is the best class to use to manage such time series
#
On Thu, 4 Oct 2012, Poizot Emmanuel wrote:

            
See the "Time Series" task view for guidance:
http://CRAN.R-project.org/view=TimeSeries
#
On Thu, Oct 4, 2012 at 3:07 AM, Hasan Diwan <hasan.diwan at gmail.com> wrote:
If you are using xts and reading in the data from an external file
then note that xts loads the zoo package and read.zoo can be used to
do the actually reading:

Lines <- "Date    Heure Profondeur Temp?rature Salinit? Turbidit? Chloration
1 2012-07-06 08:47:22     -0.144      22.469    0.011 0.000          0
2 2012-07-06 09:02:21     -0.147      22.476    0.011 0.000          0
3 2012-07-06 09:17:21     -0.139      22.498    0.011 19.323          0
4 2012-07-06 09:32:21     -0.136      22.540    0.011 19.343          0
5 2012-07-06 09:47:21     -0.141      22.510    0.011 19.321          0
6 2012-07-06 10:02:21     -0.139      22.372    0.011 19.280          0"

library(xts) # also pulls in zoo

# z <- read.zoo("myfile.dat", header = TRUE, index = 1:2, tz = "")
z <- read.zoo(text = Lines, header = TRUE, index = 1:2, tz = "")

x <- as.xts(z)

Here index = 1:2 says that the date/time index is in the first two
columns and tz = "" says to interpret it as POSIXct with the indicated
time zone.  tz = "GMT" is another possibility.

For more info on read.zoo see ?read.zoo .  Also there is a document
entirely devoted to read.zoo examples obtained by issuing:
vignette("zoo-read") .
#
Hi,
library(xts)

xts(dat1[,-1:2],order.by=as.POSIXct(paste(dat1[,1],dat1[,2],sep=" "),format="%Y-%m-%d %H:%M:%S"))
#Error in .subset(x, j) : only 0's may be mixed with negative subscripts


xts(dat1[,3:7],order.by=as.POSIXct(paste(dat1[,1],dat1[,2],sep=" "),format="%Y-%m-%d %H:%M:%S"))
#??????????????????? Profondeur Temp?rature Salinit? Turbidit? Chloration
#2012-07-06 00:00:22???? -0.145????? 22.468??? 0.011???? 0.000????????? 0
#2012-07-06 00:02:21???? -0.143????? 22.475??? 0.011???? 0.000????????? 0
#2012-07-06 01:17:21???? -0.132????? 22.456??? 0.011???? 0.323????????? 0
#2012-07-06 08:47:22???? -0.144????? 22.469??? 0.011???? 0.000????????? 0
#2012-07-06 09:02:21???? -0.147????? 22.476??? 0.011???? 0.000????????? 0
#2012-07-06 09:17:21???? -0.139????? 22.498??? 0.011??? 19.323????????? 0
#2012-07-06 09:32:21???? -0.136????? 22.540??? 0.011??? 19.343????????? 0
#2012-07-06 09:47:21???? -0.141????? 22.510??? 0.011??? 19.321????????? 0
#2012-07-06 10:02:21???? -0.139????? 22.372??? 0.011??? 19.280????????? 0
A.K.







----- Original Message -----
From: Hasan Diwan <hasan.diwan at gmail.com>
To: Poizot Emmanuel <emmanuel.poizot at cnam.fr>
Cc: r-help at r-project.org
Sent: Thursday, October 4, 2012 3:07 AM
Subject: Re: [R] Class for time series

Mr. Emmanuel,
On 4 October 2012 02:43, Poizot Emmanuel <emmanuel.poizot at cnam.fr> wrote:

            
Use xts whenever dealing with timeseries, to construct:
xts(data.in[,-1:2], order.by=as.POSIXct(paste(data.in[,1:2])))

Also, when you post help requests, use dput on your data set. I'm assuming
it's a data.frame, but I'd be able to actually test my code if there were
dput output in your question. -- H