orsf_summarize_uni: Univariate summary

Description

Summarize the univariate information from an ORSF object

Usage

orsf_summarize_uni(
  object,
  n_variables = NULL,
  pred_horizon = NULL,
  pred_type = NULL,
  importance = NULL,
  class = NULL,
  verbose_progress = FALSE,
  ...
)

Value

an object of class 'orsf_summary', which includes data on

importance of individual predictors.
expected values of predictions at specific values of predictors.

Arguments

object

(ObliqueForest) a trained oblique random forest object (see orsf).

n_variables

(integer) how many variables should be summarized? Setting this input to a lower number will reduce computation time.

pred_horizon

(double) Only relevent for survival forests. A value or vector indicating the time(s) that predictions will be calibrated to. E.g., if you were predicting risk of incident heart failure within the next 10 years, then pred_horizon = 10. pred_horizon can be NULL if pred_type is 'mort', since mortality predictions are aggregated over all event times

pred_type

(character) the type of predictions to compute. Valid Valid options for survival are:

'risk' : probability of having an event at or before pred_horizon.
'surv' : 1 - risk.
'chf': cumulative hazard function
'mort': mortality prediction
'time': survival time prediction

For classification:

'prob': probability for each class

For regression:

'mean': predicted mean, i.e., the expected value

importance

(character) Indicate method for variable importance:

'none': no variable importance is computed.
'anova': compute analysis of variance (ANOVA) importance
'negate': compute negation importance
'permute': compute permutation importance

class

(character) only relevant for classification forests. If NULL (the default), summary statistics are returned for all classes in the outcome, and printed summaries will show the last class in the class levels. To specify a single class to summarize, indicate the name of the class with class. E.g., if the categorical outcome has class levels A, B, and C, then using class = "A" will restrict output to class A.

For details on these methods, see orsf_vi.

verbose_progress

(logical) if TRUE, progress will be printed to console. If FALSE (the default), nothing will be printed.

...

Further arguments passed to or from other methods (not currently used).

Details

If pred_horizon is left unspecified, the median value of the time-to-event variable in object's training data will be used. It is recommended to always specify your own prediction horizon, as the median time may not be an especially meaningful horizon to compute predicted risk values at.

If object already has variable importance values, you can safely bypass the computation of variable importance in this function by setting importance = 'none'.

Examples

Run this code


object <- orsf(pbc_orsf, Surv(time, status) ~ . - id, n_tree = 25)

# since anova importance was used to make object, it is also
# used for ranking variables in the summary, unless we specify
# a different type of importance

orsf_summarize_uni(object, n_variables = 2)

# if we want to summarize object according to variables
# ranked by negation importance, we can compute negation
# importance within orsf_summarize_uni() as follows:

orsf_summarize_uni(object, n_variables = 2, importance = 'negate')

# for multi-category fits, you can specify which class
# you want to summarize:

object =  orsf(species ~ ., data = penguins_orsf, n_tree = 25)

orsf_summarize_uni(object, class = "Adelie", n_variables = 1)