Balanced accuracy is computed here as the average of sens() and spec().

bal_accuracy(data, ...)

# S3 method for data.frame
bal_accuracy(data, truth, estimate, estimator = NULL, na_rm = TRUE, ...)

bal_accuracy_vec(truth, estimate, estimator = NULL, na_rm = TRUE, ...)



Either a data.frame containing the truth and estimate columns, or a table/matrix where the true class results should be in the columns of the table.


Not currently used.


The column identifier for the true class results (that is a factor). This should be an unquoted column name although this argument is passed by expression and supports quasiquotation (you can unquote column names). For _vec() functions, a factor vector.


The column identifier for the predicted class results (that is also factor). As with truth this can be specified different ways but the primary method is to use an unquoted variable name. For _vec() functions, a factor vector.


One of: "binary", "macro", "macro_weighted", or "micro" to specify the type of averaging to be done. "binary" is only relevant for the two class case. The other three are general methods for calculating multiclass metrics. The default will automatically choose "binary" or "macro" based on estimate.


A logical value indicating whether NA values should be stripped before the computation proceeds.


A tibble with columns .metric, .estimator, and .estimate and 1 row of values.

For grouped data frames, the number of rows returned will be the same as the number of groups.

For bal_accuracy_vec(), a single numeric value (or NA).

Relevant Level

There is no common convention on which factor level should automatically be considered the "event" or "positive" result. In yardstick, the default is to use the first level. To change this, a global option called yardstick.event_first is set to TRUE when the package is loaded. This can be changed to FALSE if the last level of the factor is considered the level of interest by running: options(yardstick.event_first = FALSE). For multiclass extensions involving one-vs-all comparisons (such as macro averaging), this option is ignored and the "one" level is always the relevant result.


Macro, micro, and macro-weighted averaging is available for this metric. The default is to select macro averaging if a truth factor with more than 2 levels is provided. Otherwise, a standard binary calculation is done. See vignette("multiclass", "yardstick") for more information.

See also

Other class metrics: accuracy(), detection_prevalence(), f_meas(), j_index(), kap(), mcc(), npv(), ppv(), precision(), recall(), sens(), spec()


# Two class data("two_class_example") bal_accuracy(two_class_example, truth, predicted)
#> # A tibble: 1 x 3 #> .metric .estimator .estimate #> <chr> <chr> <dbl> #> 1 bal_accuracy binary 0.837
# Multiclass library(dplyr) data(hpc_cv) hpc_cv %>% filter(Resample == "Fold01") %>% bal_accuracy(obs, pred)
#> # A tibble: 1 x 3 #> .metric .estimator .estimate #> <chr> <chr> <dbl> #> 1 bal_accuracy macro 0.717
# Groups are respected hpc_cv %>% group_by(Resample) %>% bal_accuracy(obs, pred)
#> # A tibble: 10 x 4 #> Resample .metric .estimator .estimate #> <chr> <chr> <chr> <dbl> #> 1 Fold01 bal_accuracy macro 0.717 #> 2 Fold02 bal_accuracy macro 0.711 #> 3 Fold03 bal_accuracy macro 0.767 #> 4 Fold04 bal_accuracy macro 0.724 #> 5 Fold05 bal_accuracy macro 0.715 #> 6 Fold06 bal_accuracy macro 0.707 #> 7 Fold07 bal_accuracy macro 0.699 #> 8 Fold08 bal_accuracy macro 0.734 #> 9 Fold09 bal_accuracy macro 0.717 #> 10 Fold10 bal_accuracy macro 0.706
# Weighted macro averaging hpc_cv %>% group_by(Resample) %>% bal_accuracy(obs, pred, estimator = "macro_weighted")
#> # A tibble: 10 x 4 #> Resample .metric .estimator .estimate #> <chr> <chr> <chr> <dbl> #> 1 Fold01 bal_accuracy macro_weighted 0.771 #> 2 Fold02 bal_accuracy macro_weighted 0.763 #> 3 Fold03 bal_accuracy macro_weighted 0.799 #> 4 Fold04 bal_accuracy macro_weighted 0.758 #> 5 Fold05 bal_accuracy macro_weighted 0.762 #> 6 Fold06 bal_accuracy macro_weighted 0.746 #> 7 Fold07 bal_accuracy macro_weighted 0.733 #> 8 Fold08 bal_accuracy macro_weighted 0.768 #> 9 Fold09 bal_accuracy macro_weighted 0.734 #> 10 Fold10 bal_accuracy macro_weighted 0.750
# Vector version bal_accuracy_vec(two_class_example$truth, two_class_example$predicted)
#> [1] 0.8366167
# Making Class2 the "relevant" level options(yardstick.event_first = FALSE) bal_accuracy_vec(two_class_example$truth, two_class_example$predicted)
#> [1] 0.8366167
options(yardstick.event_first = TRUE)