ROC (Receiver Operating Characteristic) curve data from a classification forest.

A classifier does not hand you a class; it hands you a predicted probability, and you pick a threshold. Slide that threshold from 0 to 1 and the trade-off between catching the positives and crying wolf shifts the whole way. The ROC curve traces that trade-off. For one class of a classification rfsrc or randomForest forest, gg_roc walks every threshold and records sensitivity (the true positive rate) against specificity (1 minus the false positive rate).

Usage

# S3 method for class 'rfsrc'
gg_roc(object, which_outcome, oob = TRUE, per_class = FALSE, ...)

gg_roc(object, which_outcome, oob = TRUE, per_class = FALSE, ...)

Arguments

object

A classification rfsrc or randomForest object. Only forests with family == "class" (rfsrc) or type == "classification" (randomForest) are supported.

which_outcome

Index of the class to score: 1 or 2 for a binary forest, any valid class index for a multi-class one. A character level name is accepted by the randomForest method only; the rfsrc method requires an integer index. which_outcome = "all" or 0 behaves differently by engine:

randomForest method: Returns a macro-averaged one-vs-rest ROC computed over the per-class probabilities.
rfsrc method: Warns and falls back to class 1. The macro-average and per-class faceting for the rfsrc path are tracked separately under issue #72.

oob

Logical; if TRUE (default), build the curve from out-of-bag predicted probabilities, otherwise from full in-bag predictions. For randomForest, TRUE uses the out-of-bag vote probabilities in object$votes; FALSE uses in-bag predict(type = "prob").

per_class

Logical; if TRUE and the forest has more than two classes, return one ROC curve per class, each class scored against all the others. The result is a long-format data.frame with a class factor column and a named AUC vector attribute, ordered by descending AUC. Binary forests treat per_class = TRUE as a no-op. Honored by the randomForest method only.

...

Extra arguments (currently unused).

Value

A gg_roc data.frame, one row per unique prediction threshold, with columns:

sens: Sensitivity (true positive rate) at the threshold.
spec: Specificity (true negative rate) at the threshold.
pct: The probability threshold used for that row.

Pass it to calc_auc for the area under the curve.

Defaults across entry points

plot(gg_roc(x)) and a direct plot.gg_roc(x) call on a raw multi-class forest do not draw the same figure: gg_roc(x) returns a single curve, while plot.gg_roc given a raw forest overlays one curve per class. Plain plot(x) on a forest reaches neither, as it dispatches to the forest's own method in randomForestSRC or randomForest. Pass which_outcome explicitly when the distinction matters; issue #72 tracks reconciling the two.

Examples

## ------------------------------------------------------------
## classification example
## ------------------------------------------------------------
## -------- iris data
rfsrc_iris <- randomForestSRC::rfsrc(Species ~ ., data = iris, ntree = 100)

# ROC for setosa
gg_dta <- gg_roc(rfsrc_iris, which_outcome = 1)
plot(gg_dta)


# ROC for versicolor
gg_dta <- gg_roc(rfsrc_iris, which_outcome = 2)
plot(gg_dta)


# ROC for virginica
gg_dta <- gg_roc(rfsrc_iris, which_outcome = 3)
plot(gg_dta)


## -------- iris data
rf_iris <- randomForest::randomForest(Species ~ ., data = iris)

# ROC for setosa
gg_dta <- gg_roc(rf_iris, which_outcome = 1)
plot(gg_dta)


# ROC for versicolor
gg_dta <- gg_roc(rf_iris, which_outcome = 2)
plot(gg_dta)


# ROC for virginica
gg_dta <- gg_roc(rf_iris, which_outcome = 3)
plot(gg_dta)