MachineShop (version 2.0.0)

performance: Model Performance Metrics

Description

Compute measures of model performance.

Usage

performance(x, ...)
# S3 method for BinomialVariate
performance(
  x,
  y,
  metrics = MachineShop::settings("metrics.numeric"),
  na.rm = TRUE,
  ...
)
# S3 method for factor
performance(
  x,
  y,
  metrics = MachineShop::settings("metrics.factor"),
  cutoff = MachineShop::settings("cutoff"),
  na.rm = TRUE,
  ...
)
# S3 method for matrix
performance(
  x,
  y,
  metrics = MachineShop::settings("metrics.matrix"),
  na.rm = TRUE,
  ...
)
# S3 method for numeric
performance(
  x,
  y,
  metrics = MachineShop::settings("metrics.numeric"),
  na.rm = TRUE,
  ...
)
# S3 method for Surv
performance(
  x,
  y,
  metrics = MachineShop::settings("metrics.Surv"),
  cutoff = MachineShop::settings("cutoff"),
  na.rm = TRUE,
  ...
)
# S3 method for ConfusionList
performance(x, ...)
# S3 method for ConfusionMatrix
performance(x, metrics = MachineShop::settings("metrics.ConfusionMatrix"), ...)
# S3 method for Resamples
performance(x, ...)

Arguments

x

observed responses; or confusion or resample result containing observed and predicted responses.

...

arguments passed from the Resamples method to the response type-specific methods or from the method for ConfusionList to ConfusionMatrix.

y

predicted responses if not contained in x.

metrics

metric function, function name, or vector of these with which to calculate performance.

na.rm

logical indicating whether to remove observed or predicted responses that are NA when calculating metrics.

cutoff

numeric (0, 1) threshold above which binary factor probabilities are classified as events and below which survival probabilities are classified.

See Also

plot, summary

Examples

# NOT RUN {
res <- resample(Species ~ ., data = iris, model = GBMModel)
(perf <- performance(res))
summary(perf)
plot(perf)

## Survival response example
library(survival)
library(MASS)

fo <- Surv(time, status != 2) ~ sex + age + year + thickness + ulcer
gbm_fit <- fit(fo, data = Melanoma, model = GBMModel)

obs <- response(gbm_fit, newdata = Melanoma)
pred <- predict(gbm_fit, newdata = Melanoma, type = "prob")
performance(obs, pred)

# }