smooth_ts: Smooth time series

In placeboo/amplify: Forecasting Daily Time Seriers with Double Seasonal Patterns

Description

Smooth the time series by impute the outlers.

Usage

smooth_ts(
  data,
  vars = c("date", "ts"),
  par = list(lambda = 1.5, s = 7),
  weekday.sep = FALSE
)

Arguments

data	A data frame includes dates and time series columns
vars	A character vector with variable names with respect to date and time series
par	A list of parameters. lambda = 1.5, the "distance" describes the outlier and the main population. s = 7, the most obvious seasonality of the time series is 7 (days).
weekday.sep	A logical value indicating whether model each weekday time series separately.

Details

Use find_outlier to find the outliers and use one seasonal exponential smoothing to imput outliers

Value

A list with:

• ts.dat. A data frame with column name is_outlier and ts. is_outlier is a character vector that indicates whether the observation is outier ("yes", "no"). ts is a numeric vector that stores the time series after smoothing.

• model. An object of class "ets", see ets.

See Also

find_outlier

Examples

data(tickets)
smooth.dat = smooth_ts(tickets, vars = c("date", "ticket"))
cbind(tickets, smooth.dat)

placeboo/amplify documentation built on Oct. 6, 2020, 9:04 a.m.