Rolling-origin evaluation for time-aware Gaussian processes

Performs expanding-window rolling-origin evaluation with fixed kernel and noise hyperparameters. Each split is fitted using only observations available at that origin, preventing look-ahead leakage.

Usage

rolling_origin_gp(
  time,
  y,
  kernel,
  initial_window,
  horizon = 1L,
  step = 1L,
  noise_variance = 1e-06,
  mean = zero_mean(),
  unit = "auto",
  interval_level = 0.95,
  ...
)

Arguments

time

Numeric, Date, or POSIXt vector in strictly increasing order.

y

Numeric response vector.

kernel

Fixed Gaussian-process kernel specification.

initial_window

Number of observations in the first training window.

horizon

Number of future observations predicted at each origin.

step

Number of observations between consecutive origins.

noise_variance

Non-negative observation-noise variance: a scalar, or one value per observation. With one value per observation, each training window uses the values of its rows and each forecast uses the values of its target rows.

mean

Gaussian-process mean specification.

unit

Time unit passed to gp_time_index().

interval_level

Probability level for prediction intervals.

...

Additional arguments passed to fit_time_series_gp(), such as method = "state_space".

Value

A data frame with class gaussianprocesses_backtest.

Details

Fixed hyperparameters prevent future observations from influencing earlier origins through repeated whole-series hyperparameter tuning. Hyperparameters can be selected separately using an appropriate training period.

Stability

Stable: from version 1.0.0 this interface changes incompatibly only in a major release, after a deprecation period. Results and options that concern an experimental model class, kernel, or argument follow that interface's tier. See gaussianprocesses-package for the policy.

Examples

time <- 0:23
y <- sin(2 * pi * time / 12)

backtest <- rolling_origin_gp(
  time,
  y,
  kernel = time_series_kernel(period = 12),
  initial_window = 12,
  horizon = 3,
  step = 3,
  noise_variance = 0.05
)
head(backtest[, c("origin_index", "horizon", "observed", "mean", "covered")])
#> Gaussian-process rolling-origin backtest
#>   forecasts: 6
#>   origins: 2
#>   maximum horizon: 3