Package 'tframePlus'

Aggregate Time Series

Description

Convert series to quarterly or annual.

Usage

as.quarterly(x, FUN=sum, na.rm=FALSE, ...)
    as.annually(x, FUN=sum, na.rm=FALSE, ...)
    as.weekly(x, FUN=sum, na.rm=FALSE, foldFrom=end(x), periodicity = 7)

Arguments

x	a tframed object. (Only montly is currently working)
FUN	the function to use for aggregating.
na.rm	Logical indicating if NA should be removed from the beginning and end of a series
foldFrom	a date which is used to determine the end of weeks.
periodicity	the number of periods in a week.
...	additional arguments passed to aggregate

Details

Functions as.quarterly and as.annually uses aggregate, but shifts the data to match usual economic and financial quarters or years (whereas aggregate simply groups together a number of periods corresponding to the new frequency starting with the first observation). These functions should give the same result as aggregate if the data starts and ends on quarter (annual) boundaries. If the data does not start and end on quarter (annual) boundaries then NA will be put in where data is incomplete, and the quarter (year) removed if na.rm=TRUE.

Monthly to annual gives the aggregate by converting first to quarterly, which is not exactly correct.

To weekly by periodicity groupings from foldFrom. Partial weeks in the beginning or end are padded with NA and dropped if na.rm=FALSE. (as.weekly has been tested only with daily 7 day weeks.)

Value

time series

See Also

tfwindow aggregate

Examples

z <- ts(1:10, start = c(1999,2), frequency=4)
  as.annually(z) 
  as.annually(z, na.rm=TRUE)

  z <- ts(1:30, start = c(1999,2), frequency=12)
  as.annually(z) 
  as.annually(z, na.rm=TRUE)
  as.quarterly(z) 
  as.quarterly(z, na.rm=TRUE)

---

Change the class of a (Multivariate) Time Series

Description

Change the class of a (Multivariate) Time Series

Usage

changeTSrepresentation (x, newRepresentation)

Arguments

x	a multivariate time series object.
newRepresentation	function or character string indicating the function to use to change the time series class.

Details

This function attempts to convert the time series to the new class using newRepresentation.

If the representation is a character string then it is applied using do.call(TSrepresentation, list(x, dates)) where x is the time series (matrix) to be return and dates are determined by as.Date(time(x)) where x is the default representation of the data. If representation is not a character string then it should be a function and is applied using TSrepresentation(x, dates)

Value

a multivariate time series object.

---

Rolling Time Series Aggregate

Description

Aggregate time series using a rolling window.

Usage

rollAggregate(x, FUN=sum, na.rm=FALSE, aggPeriods=4, ...)

Arguments

x	a time series object.
FUN	the function to use for aggregating.
na.rm	Logical indicating if NA should be removed from the beginning and end of a series
aggPeriods	the number of observations to use in the rolling window.
...	additional arguments passed to apply

Details

Functions rollAggregate uses apply with FUN on a rolling window set by aggPeriods. Partial windows may result in NA, depending on FUN. NA in the beginning or end are dropped if na.rm=TRUE. The frequency of the resulting series remains the same as the original, but the result will be aggPeriods shorter (i.e. have NAs). The default will be a four period moving sum (which roughly converts weekly data to something approximating monthly, but at a weekly frequency).

Value

time series

See Also

as.annually, as.quarterly, as.weekly, apply

Examples

rollAggregate(ts(1:20, start = c(1999,1), frequency=1)) 
  rollAggregate(ts(1:20, start = c(1999,1), frequency=12), aggPeriods=3) 
  
  if(require("zoo")) {
    z <- zoo(rnorm(100),  order.by = Sys.Date() + 1:100)
    rollAggregate(as.weekly(z), aggPeriods=4, FUN=mean)
    require("tfplot")
    tfplot(as.weekly(z),rollAggregate(as.weekly(z), aggPeriods=4, 
       FUN=mean, na.rm=TRUE))
    tfplot(z,rollAggregate(z, aggPeriods=28, FUN=mean))
    }

---

Integrate Multivariate Time Series

Description

Integrate multivariate time series using cumsum

Usage

tfI (x, I=rep(TRUE, NCOL(x)), t0=rep(0, NCOL(x)))

Arguments

x	a multivariate time series object.
I	a logical of length equal the number of series in x indicating which series to integrate, default all.
t0	a vector of length equal the number of series in x indicating initial (time zero) valuess, defaults to zero.

Details

This function first prepends t0 to x and then applies cumsum to each series indicated by I. The first value is then dropped from the result.

Value

a multivariate time series object of the same dimension as x.

See Also

tfL tfwindow aggregate cumsum as.quarterly

---

Plot Series With persp

Description

Plot series with persp

Usage

tfpersp (x, tf=tfspan(x), start=tfstart(tf), end=tfend(tf),
       theta = -30, phi = 15, scale = FALSE, 
       xlab = "Time", ylab = "", zlab = "", 
       aspect= c(0.5, 0.5), #y/time, z/time,
       ticktype="detailed",ltheta = -120, lphi = 15,
       ...)

Arguments

x	multivariate time series (a tframed object).
tf	a tframe to set the start and end of the plot.
start	start period for the plot.
end	end period for the plot.
theta	argument passed to persp
phi	argument passed to persp
scale	argument passed to persp
xlab	argument passed to persp
ylab	argument passed to persp
zlab	argument passed to persp
aspect	argument passed to persp
ticktype	argument passed to persp
ltheta	argument passed to persp
lphi	argument passed to persp
...	additional arguments passed to persp

Details

This function uses passes a multivariate time series to persp.

Value

Depends

See Also

tfwindow persp

---

Write Time Series to an .xls Spreadsheet File

Description

Write one or more time series to a .xls spreadsheet file

Usage

TSwriteXLS(x, ..., FileName="R.xls", SheetNames=NULL,
               dateHeader="date", verbose = FALSE)
    TSwriteCSV(x, FileName="R.csv",  dateFormat=1, dateHeader="date")

Arguments

x	a time series or matrix of time series.)
...	additional time series objects (like x).
FileName	a string to use for the file name.
SheetNames	Logical indicating if NA should be removed from the beginning and end of a series.
dateFormat	an integer indicating the format to use for dates. (See details.)
dateHeader	string or vector of strings to use as header on date column.
verbose	logical passed to WriteXLS.

Details

This functions uses WriteXLS to write time series data to an .xls spreadsheet file. If the object x has mulitple series then they will be put on the same sheet. Additional objects in ... will be put on additional pages.

The first line of each sheet will have column headers using seriesNames from the series in the column. The first column will have dates and the heading for that column will be determined by dateHeader. This can be a vector of length equal the number of sheets, of a single character string, in which case it is replicated for each sheet. The dateHeader is a might be useful for indication information like the date of the last observation in a known place on the sheet (A1) for indicating on a graph. For convenience, the year, period, and a character representation of the period are indicated in columns 2 to 4. This is followed by columns of the data.

If SheetNames is NULL (the default) then the sheet names will be generated from the first series name in each object. If SheetNames is supplied then it should have an element for x and for each object in ....

If WriteXLS does not find a version of perl with appropriate modules then a work around is attempted using save.

WriteCSV handles only a single object x which can have mulitple series. These are wtten as columns in the comma separated value file. dateFormat can be 0 indicating no date, 1 indicating a format like "Jan 1969", 2 indicating a format like 1969,1, or 3 indicating a format like 1969,"Jan",. Only 0 or 1 are supported for data other than monthly and quarterly frequencies.

Value

logical

See Also

WriteXLS

Examples

tofile <- tempfile()
 
 z <- ts(1:10, start=c(1999,2), freq=12)
 seriesNames(z) <- "ser 1"
 TSwriteXLS(z, FileName=tofile) 

 zz <- tbind(z, diff(z))
 seriesNames(zz) <- c("ser 1", "diff")
 TSwriteXLS(zz, FileName=tofile,  SheetNames="2 series")

 zz <- ts(1:10, start=c(1999,1), freq=1)
 seriesNames(zz) <- "annual ser"
 TSwriteXLS(z, zz, FileName=tofile,  SheetNames=c("monthly", "annual"))

 unlink(tofile)

---