Summarises which metrics were selected and what happened to the rest. Prints a summary rather than the ranking table itself, which has one row per metric passed in and is usually too long to read in the console; use `x$ranking` to see it.
Usage
# S3 method for class 'metrics_evaluation'
print(x, ...)Arguments
- x
A `metrics_evaluation` object from
evaluate_metrics.- ...
Ignored, for compatibility with the generic.
See also
Other metrics:
calculate_metrics(),
evaluate_metrics()
Examples
landscapes <- create_landscapes(n = 10, patterns = c("spots", "random"))
#> ✔ Successfully generated all 10 training landscapes
metrics <- calculate_metrics(landscapes, level = "landscape")
#> ■■■■■■■■■■■■■ 39% | ETA: 6s
#> ■■■■■■■■■■■■■■■■■■■■■ 67% | ETA: 3s
evaluate_metrics(metrics, metrics_number = 5)
#> Warning: Excluded 6 metrics with missing values (60 rows removed).
#> ✖ NA value for at least one landscape: "enn_cv", "enn_mn", "enn_sd", "iji",
#> "pafrac", and "rpr"
#> ℹ Use `exclude_incomplete_metrics = FALSE` to retain them (not recommended for
#> model training).
#> Warning: Excluded 3 metrics with no variation across landscapes: "pr", "prd", and "ta"
#> Metrics evaluation: kruskal_effsize [66 candidate metrics]
#> -----------------------------------------
#> Selected (5): area_mn, dcore_sd, para_cv, circle_cv, division
#>
#> Outcomes:
#> selected 5
#> dropped_correlated 37
#> dropped_below_cutoff 15
#> excluded_incomplete 6
#> excluded_zero_variance 3
#>
#> Use $ranking for scores and per-metric outcomes.