rsample (version 0.1.0)

rolling_origin: Rolling Origin Forecast Resampling

Description

This resampling method is useful when the data set has a strong time component. The resamples are not random and contain data points that are consecutive values. The function assumes that the original data set are sorted in time order.

Usage

rolling_origin(
  data,
  initial = 5,
  assess = 1,
  cumulative = TRUE,
  skip = 0,
  lag = 0,
  ...
)

Arguments

data

A data frame.

initial

The number of samples used for analysis/modeling in the initial resample.

assess

The number of samples used for each assessment resample.

cumulative

A logical. Should the analysis resample grow beyond the size specified by initial at each resample?.

skip

A integer indicating how many (if any) additional resamples to skip to thin the total amount of data points in the analysis resample. See the example below.

lag

A value to include a lag between the assessment and analysis set. This is useful if lagged predictors will be used during training and testing.

...

Not currently used.

Value

An tibble with classes rolling_origin, rset, tbl_df, tbl, and data.frame. The results include a column for the data split objects and a column called id that has a character string with the resample identifier.

Details

The main options, initial and assess, control the number of data points from the original data that are in the analysis and assessment set, respectively. When cumulative = TRUE, the analysis set will grow as resampling continues while the assessment set size will always remain static. skip enables the function to not use every data point in the resamples. When skip = 0, the resampling data sets will increment by one position. Suppose that the rows of a data set are consecutive days. Using skip = 6 will make the analysis data set to operate on weeks instead of days. The assessment set size is not affected by this option.

See Also

sliding_window(), sliding_index(), and sliding_period() for additional time based resampling functions.

Examples

# NOT RUN {
set.seed(1131)
ex_data <- data.frame(row = 1:20, some_var = rnorm(20))
dim(rolling_origin(ex_data))
dim(rolling_origin(ex_data, skip = 2))
dim(rolling_origin(ex_data, skip = 2, cumulative = FALSE))

# You can also roll over calendar periods by first nesting by that period,
# which is especially useful for irregular series where a fixed window
# is not useful. This example slides over 5 years at a time.
library(dplyr)
library(tidyr)
data(drinks, package = "modeldata")

drinks_annual <- drinks %>%
  mutate(year = as.POSIXlt(date)$year + 1900) %>%
  nest(-year)

multi_year_roll <- rolling_origin(drinks_annual, cumulative = FALSE)

analysis(multi_year_roll$splits[[1]])
assessment(multi_year_roll$splits[[1]])

# }