Skip to contents

Generate Predictions

Source: R/fitter.R
predict_fit.Rd

Generate predictions from a fitted model.

Usage

predict_fit(fitter, fit_result, newdata = NULL, seed = NULL)

Arguments

fitter

An S7 Fitter object

fit_result

A bayesim_fit_result object from fit_model()

newdata

Data frame with new observations for prediction. If NULL, predictions are generated for the original training data.

seed

Optional integer seed for reproducible predictions

Value

A list containing:

  • predicted_mean: Vector of mean predictions (length N)

  • predicted_samples: Matrix of posterior predictive samples (S x N; draws as rows, observations as columns)

  • predicted_sd: Vector of prediction standard deviations (length N)

  • Additional fitter-specific outputs

predicted_samples follows the same orientation convention as log_lik_matrix() and predict_epred(): all matrices are draws x observations (S rows, N columns).