qerror: Estimated conditional prediction error quantile functions
In forestError: A Unified Framework for Random Forest Prediction Error Estimation

Description Usage Arguments Details Value Author(s) See Also Examples

View source: R/qerror.R

Returns quantiles of the estimated conditional error distribution associated with each test prediction.

1	qerror(p, xs)

`p`	A vector of probabilities.
`xs`	A vector of the indices of the test observations for which the conditional error quantiles are desired. Defaults to all test observations given in the call of `quantForestError`.

This function is only defined as output of the quantForestError function. It is not exported as a standalone function. See the example.

If either p or xs has length one, then a vector is returned with the desired quantiles. If both have length greater than one, then a data.frame of quantiles is returned, with rows corresponding to the inputted xs and columns corresponding to the inputted p.

Benjamin Lu <b.lu@berkeley.edu>; Johanna Hardin <jo.hardin@pomona.edu>

quantForestError

# load data
data(airquality)

# remove observations with missing predictor variable values
airquality <- airquality[complete.cases(airquality), ]

# get number of observations and the response column index
n <- nrow(airquality)
response.col <- 1

# split data into training and test sets
train.ind <- sample(1:n, n * 0.9, replace = FALSE)
Xtrain <- airquality[train.ind, -response.col]
Ytrain <- airquality[train.ind, response.col]
Xtest <- airquality[-train.ind, -response.col]
Ytest <- airquality[-train.ind, response.col]

# fit random forest to the training data
rf <- randomForest::randomForest(Xtrain, Ytrain, nodesize = 5,
                                 ntree = 500,
                                 keep.inbag = TRUE)

# estimate conditional error distribution functions
output <- quantForestError(rf, Xtrain, Xtest,
                           what = c("p.error", "q.error"))

# get the 0.25 and 0.8 quantiles of the error distribution
# associated with each test prediction
output$qerror(c(0.25, 0.8))

# same as above but only for the first three test observations
output$qerror(c(0.25, 0.8), 1:3)