HarveyCumulator: HarveyCumulator

In bsts: Bayesian Structural Time Series

HarveyCumulator

Description

Given a state space model on a fine scale, the Harvey cumulator aggregates the model to a coarser scale (e.g. from days to weeks, or weeks to months).

Usage

HarveyCumulator(fine.series,
                contains.end,
                membership.fraction)

Arguments

fine.series	The fine-scale time series to be aggregated.
contains.end	A logical vector, with length matching fine.series indicating whether each fine scale time interval contains the end of a coarse time interval. For example, months don't contain a fixed number of weeks, so when cumulating a weekly time series into a monthly series, you need to know which weeks contain the end of a month.
membership.fraction	The fraction of each fine-scale time observation belonging to the coarse scale time observation at the beginning of the time interval. For example, if week i started in March and ended in April, membership.fraction[i] is the fraction of fine.series[i] that should be attributed to March. This should be 1 for most observations.

Value

Returns a vector containing the course scale partial aggregates of fine.series.

Author(s)

Steven L. Scott [e-mail]

References

Harvey (1990), "Forecasting, structural time series, and the Kalman filter", Cambridge University Press.

Durbin and Koopman (2001), "Time series analysis by state space methods", Oxford University Press.

See Also

bsts.mixed,

Examples

data(goog)
days <- factor(weekdays(index(goog)),
               levels = c("Monday", "Tuesday", "Wednesday",
                          "Thursday", "Friday"),
               ordered = TRUE)

## Because of holidays, etc the days do not always go in sequence.
## (Sorry, Rebecca Black! https://www.youtube.com/watch?v=kfVsfOSbJY0)
## diff.days[i] is the number of days between days[i-1] and days[i].
## We know that days[i] is the end of a week if diff.days[i] < 0.
diff.days <- tail(as.numeric(days), -1) - head(as.numeric(days), -1)
contains.end <- c(FALSE, diff.days < 0)

goog.weekly <- HarveyCumulator(goog, contains.end, 1)

bsts documentation built on May 29, 2024, 2:14 a.m.