These functions calculate the spec() (specificity) of a measurement system compared to a reference result (the "truth" or gold standard). Highly related functions are sens(), ppv(), and npv().

spec(data, ...)

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

specificity(data, ...)

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

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

Arguments

data

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.

truth

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.

estimate

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.

estimator

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.

na_rm

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

Value

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 spec_vec(), a single numeric value (or NA).

Details

The specificity measures the proportion of negatives that are correctly identified as negatives.

When the denominator of the calculation is 0, specificity is undefined. This happens when both # true_negative = 0 and # false_positive = 0 are true, which mean that there were no true negatives. When computing binary specificity, a NA value will be returned with a warning. When computing multiclass specificity, the individual NA values will be removed, and the computation will procede, with a warning.

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.

Multiclass

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.

Implementation

Suppose a 2x2 table with notation:

Reference
PredictedPositiveNegative
PositiveAB
NegativeCD

The formulas used here are:

$$Sensitivity = A/(A+C)$$ $$Specificity = D/(B+D)$$ $$Prevalence = (A+C)/(A+B+C+D)$$ $$PPV = (Sensitivity * Prevalence) / ((Sensitivity * Prevalence) + ((1-Specificity) * (1-Prevalence)))$$ $$NPV = (Specificity * (1-Prevalence)) / (((1-Sensitivity) * Prevalence) + ((Specificity) * (1-Prevalence)))$$

See the references for discussions of the statistics.

References

Altman, D.G., Bland, J.M. (1994) ``Diagnostic tests 1: sensitivity and specificity,'' British Medical Journal, vol 308, 1552.

See also

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

Other sensitivity metrics: npv(), ppv(), sens()

Examples

# Two class data("two_class_example") spec(two_class_example, truth, predicted)
#> # A tibble: 1 x 3 #> .metric .estimator .estimate #> <chr> <chr> <dbl> #> 1 spec binary 0.793
# Multiclass library(dplyr) data(hpc_cv) hpc_cv %>% filter(Resample == "Fold01") %>% spec(obs, pred)
#> # A tibble: 1 x 3 #> .metric .estimator .estimate #> <chr> <chr> <dbl> #> 1 spec macro 0.886
# Groups are respected hpc_cv %>% group_by(Resample) %>% spec(obs, pred)
#> # A tibble: 10 x 4 #> Resample .metric .estimator .estimate #> <chr> <chr> <chr> <dbl> #> 1 Fold01 spec macro 0.886 #> 2 Fold02 spec macro 0.882 #> 3 Fold03 spec macro 0.899 #> 4 Fold04 spec macro 0.879 #> 5 Fold05 spec macro 0.881 #> 6 Fold06 spec macro 0.873 #> 7 Fold07 spec macro 0.866 #> 8 Fold08 spec macro 0.884 #> 9 Fold09 spec macro 0.867 #> 10 Fold10 spec macro 0.875
# Weighted macro averaging hpc_cv %>% group_by(Resample) %>% spec(obs, pred, estimator = "macro_weighted")
#> # A tibble: 10 x 4 #> Resample .metric .estimator .estimate #> <chr> <chr> <chr> <dbl> #> 1 Fold01 spec macro_weighted 0.816 #> 2 Fold02 spec macro_weighted 0.815 #> 3 Fold03 spec macro_weighted 0.839 #> 4 Fold04 spec macro_weighted 0.803 #> 5 Fold05 spec macro_weighted 0.812 #> 6 Fold06 spec macro_weighted 0.795 #> 7 Fold07 spec macro_weighted 0.790 #> 8 Fold08 spec macro_weighted 0.814 #> 9 Fold09 spec macro_weighted 0.795 #> 10 Fold10 spec macro_weighted 0.801
# Vector version spec_vec(two_class_example$truth, two_class_example$predicted)
#> [1] 0.7933884
# Making Class2 the "relevant" level options(yardstick.event_first = FALSE) spec_vec(two_class_example$truth, two_class_example$predicted)
#> [1] 0.879845
options(yardstick.event_first = TRUE)