Returns row-level selected and exploratory assignments from a fitted object.
Arguments
- fit
A fitted object from
fit_uccdf().
Value
A data frame with one row per sample. When the global null is not
rejected, cluster remains 1 while exploratory_* columns expose the best
unsupported multi-cluster split.
Examples
dat <- simulate_mixed_data(n = 40, k = 2, seed = 1)
fit <- fit_uccdf(dat, id_column = "sample_id", n_resamples = 10, n_null = 39, seed = 1)
head(augment(fit))
#> row_id cluster confidence ambiguity exploratory_cluster
#> S001 S001 1 0.9485380 5.146199e-02 1
#> S002 S002 2 1.0000000 5.593150e-10 2
#> S003 S003 2 1.0000000 5.699666e-10 2
#> S004 S004 1 0.7244848 2.755152e-01 1
#> S005 S005 2 1.0000000 5.551379e-10 2
#> S006 S006 1 0.9602339 3.976608e-02 1
#> exploratory_confidence exploratory_ambiguity assignment_mode selected_k
#> S001 0.9485380 5.146199e-02 selected 2
#> S002 1.0000000 5.593150e-10 selected 2
#> S003 1.0000000 5.699666e-10 selected 2
#> S004 0.7244848 2.755152e-01 selected 2
#> S005 1.0000000 5.551379e-10 selected 2
#> S006 0.9602339 3.976608e-02 selected 2
#> exploratory_k
#> S001 2
#> S002 2
#> S003 2
#> S004 2
#> S005 2
#> S006 2