Once desirability columns have been created, determine the overall desirability using a mean (geometric by default).
Arguments
- ...
One or more unquoted expressions separated by commas. To choose multiple columns using selectors,
dplyr::across()can be used (see the example below).- geometric
A logical for whether the geometric or arithmetic mean should be used to summarize the columns.
- tolerance
A numeric value where values strictly less than this value are capped at the value. For example, if users wish to use the geometric mean without completely excluding settings, a value greater than zero can be used.
Examples
library(dplyr)
#>
#> Attaching package: ‘dplyr’
#> The following objects are masked from ‘package:stats’:
#>
#> filter, lag
#> The following objects are masked from ‘package:base’:
#>
#> intersect, setdiff, setequal, union
# Choose model tuning parameters that minimize the number of predictors used
# while maximizing the area under the ROC curve.
classification_results %>%
mutate(
d_feat = d_min(num_features, 1, 200),
d_roc = d_max(roc_auc, 0.5, 0.9),
d_all = d_overall(across(starts_with("d_")))
) %>%
arrange(desc(d_all))
#> # A tibble: 300 × 9
#> mixture penalty mn_log_loss pr_auc roc_auc num_features d_feat d_roc
#> <dbl> <dbl> <dbl> <dbl> <dbl> <int> <dbl> <dbl>
#> 1 0.667 0.0149 0.201 0.463 0.858 14 0.935 0.896
#> 2 0.889 0.0149 0.204 0.461 0.852 11 0.950 0.880
#> 3 1 0.0149 0.205 0.462 0.850 10 0.955 0.875
#> 4 1 0.0117 0.202 0.461 0.854 12 0.945 0.884
#> 5 0.667 0.0189 0.204 0.462 0.853 12 0.945 0.883
#> 6 0.889 0.0117 0.201 0.464 0.857 14 0.935 0.893
#> 7 1 0.00924 0.200 0.465 0.859 15 0.930 0.897
#> 8 1 0.0189 0.208 0.459 0.844 7 0.970 0.859
#> 9 0.778 0.0189 0.205 0.461 0.850 11 0.950 0.876
#> 10 0.556 0.0189 0.202 0.463 0.858 15 0.930 0.894
#> # ℹ 290 more rows
#> # ℹ 1 more variable: d_all <dbl>
# Bias the ranking toward minimizing features by using a larger scale.
classification_results %>%
mutate(
d_feat = d_min(num_features, 1, 200, scale = 3),
d_roc = d_max(roc_auc, 0.5, 0.9),
d_all = d_overall(across(starts_with("d_")))
) %>%
arrange(desc(d_all))
#> # A tibble: 300 × 9
#> mixture penalty mn_log_loss pr_auc roc_auc num_features d_feat d_roc
#> <dbl> <dbl> <dbl> <dbl> <dbl> <int> <dbl> <dbl>
#> 1 1 0.0189 0.208 0.459 0.844 7 0.912 0.859
#> 2 1 0.0240 0.213 0.452 0.835 6 0.927 0.837
#> 3 0.778 0.0240 0.209 0.459 0.844 8 0.898 0.859
#> 4 0.667 0.0304 0.211 0.457 0.842 8 0.898 0.855
#> 5 0.889 0.0240 0.210 0.456 0.841 8 0.898 0.852
#> 6 0.778 0.0304 0.213 0.452 0.835 7 0.912 0.839
#> 7 0.889 0.0304 0.216 0.448 0.830 6 0.927 0.826
#> 8 0.556 0.0386 0.213 0.455 0.840 8 0.898 0.850
#> 9 1 0.0149 0.205 0.462 0.850 10 0.870 0.875
#> 10 0.667 0.0386 0.216 0.449 0.833 7 0.912 0.832
#> # ℹ 290 more rows
#> # ℹ 1 more variable: d_all <dbl>