Skip to contents

Summary Method for ensemble_pred_fit Objects

Source: R/ensemble_pred.R
summary.ensemble_pred_fit.Rd

Provides a comprehensive summary of an ensemble prediction fit, including descriptive statistics, prediction accuracy metrics, BLP calibration test, and GAVS group averages. When train_idx was used, accuracy metrics are computed only on training observations (where Y is observed).

Usage

# S3 method for class 'ensemble_pred_fit'
summary(object, n_groups = 3, ...)

Arguments

object

An object of class ensemble_pred_fit from ensemble_pred.

n_groups

Number of groups for GAVS analysis (default: 3).

...

Additional arguments (currently unused).

Value

Invisibly returns a summary.ensemble_pred_fit object containing:

  • call: The original function call

  • outcome: Name of the outcome variable

  • n: Total number of observations

  • n_train: Number of training observations

  • M: Number of repetitions

  • metrics: Prediction accuracy metrics (R-squared, RMSE, MAE, correlation)

  • blp: BLP calibration test results

  • gavs: GAVS group average results

Examples

if (FALSE) { # \dontrun{
data(microcredit)
covars <- c("age", "gender", "education", "hhinc_yrly_base",
            "css_creditscorefinal")
dat <- microcredit[, c("bank_profits_pp", covars)]
fit <- ensemble_pred(
  bank_profits_pp ~ ., data = dat,
  train_idx = microcredit$loan_size > 0 & microcredit$treat == 1,
  algorithms = c("lm", "grf"), M = 3, K = 3
)
summary(fit)
summary(fit, n_groups = 5)
} # }