Annotated observations from a fitted drift detector

With new_data = NULL, returns the accumulated history (baseline + advanced batches) annotated with .warning, .drift and .phase. With new_data, returns a READ-ONLY preview: the batch annotated from the current state, WITHOUT persisting it — use advance() to persist.

Usage

# S3 method for class 'drift_detector_fit'
augment(x, new_data = NULL, ...)

Arguments

x

A drift_detector_fit.

new_data

Optional data frame with a new batch to preview.

...

Not used.

Value

A tibble.

Examples

base <- sim_drift_stream(n_pre = 100, n_post = 0, seed = 1)
f0 <- fit(drift_detector("ddm"), base, signal = error)
augment(f0)
#> # A tibble: 100 × 6
#>        t error drift_true .warning .drift .phase  
#>    <int> <int> <lgl>      <lgl>    <lgl>  <chr>   
#>  1     1     0 FALSE      NA       NA     baseline
#>  2     2     0 FALSE      NA       NA     baseline
#>  3     3     0 FALSE      NA       NA     baseline
#>  4     4     0 FALSE      NA       NA     baseline
#>  5     5     0 FALSE      NA       NA     baseline
#>  6     6     0 FALSE      NA       NA     baseline
#>  7     7     0 FALSE      NA       NA     baseline
#>  8     8     0 FALSE      NA       NA     baseline
#>  9     9     0 FALSE      NA       NA     baseline
#> 10    10     0 FALSE      NA       NA     baseline
#> # ℹ 90 more rows