threeRuleSmooth {disaggR}  R Documentation 
threeRuleSmooth bends a time series with a time series of a lower frequency. The procedure involved is a proportional Denton benchmark.
Therefore, the resulting time series is the product of the high frequency input with a smoothed rate. This latter is extrapolated through an arithmetic sequence.
The resulting time series is equal to the lowfrequency series after aggregation within the benchmark window.
threeRuleSmooth(
hfserie,
lfserie,
start.benchmark = NULL,
end.benchmark = NULL,
start.domain = NULL,
end.domain = NULL,
start.delta.rate = NULL,
end.delta.rate = NULL,
set.delta.rate = NULL,
...
)
hfserie 
the bended time series. It can be a matrix time series. 
lfserie 
a time series whose frequency divides the frequency of

start.benchmark 
an optional start for 
end.benchmark 
an optional end for 
start.domain 
an optional start of the output highfrequency series. It
also defines the smoothing window :
The lowfrequency residuals will be extrapolated until they contain the
smallest lowfrequency window that is around the highfrequency domain
window.
Should be a numeric of length 1 or 2, like a window for 
end.domain 
an optional end of the output highfrequency series. It also defines the smoothing window : The lowfrequency residuals will be extrapolated until they contain the smallest lowfrequency window that is around the highfrequency domain window. 
start.delta.rate 
an optional start for the mean of the rate difference.
It is required as a common difference for the arithmetical extrapolation of
the rate.
Should be a numeric of length 1 or 2, like a window for 
end.delta.rate 
an optional end for the mean of the rate difference.
It is required as a common difference for the arithmetical extrapolation of
the rate.
Should be a numeric of length 1 or 2, like a window for 
set.delta.rate 
an optional double, that allows the user to set the delta mean instead of using a mean. 
... 
if the dots contain a cl item, its value overwrites the value of the returned call. This feature allows to build wrappers. 
In order to smooth the rate, threeRuleSmooth calls bflSmooth and uses a modified and extrapolated version of hfserie as weights :
only the full cycles are kept
the first and last full cycles are replicated respectively backwards and forwards to fill the domain window.
threeRuleSmooth returns an object of class "threeRuleSmooth"
.
The functions plot
and autoplot
(the generic from ggplot2) produce
graphics of the benchmarked series and the bending series.
The functions in_disaggr, in_revisions, in_scatter
produce various comparisons on which plot and autoplot can also be used.
The generic accessor functions as.ts
, model.list
, smoothed.rate
extract
various useful features of the returned value.
An object of class "threeRuleSmooth"
is a list containing the following
components :
benchmarked.serie 
a time series, that is the result of the benchmark. 
lfrate 
a time series, that is the lowfrequency rate of the threeRuleSmooth. 
smoothed.rate 
the smoothed rate of the threeRuleSmooth. 
hfserie.as.weights 
the modified and extrapolated hfserie (see details). 
delta.rate 
the lowfrequency delta of the rate, used to extrapolate the lowfrequenccy rate time series. It is estimated as the mean value in the specified window. 
model.list 
a list containing all the arguments submitted to the function. 
call 
the matched call. 
## How to use threeRuleSmooth
smooth < threeRuleSmooth(hfserie = turnover,
lfserie = construction)
as.ts(smooth)
coef(smooth)
summary(smooth)
library(ggplot2)
autoplot(in_disaggr(smooth))