Back-Calculate Sample Concentrations from Posterior Draws

For each test sample, evaluates the inverse model at every posterior draw to produce a full posterior distribution of predicted concentration.

Usage

predict_samples_bayes(
  samples,
  bayes_fit,
  curve_idx = 1L,
  response_variable,
  is_log_response = TRUE,
  n_draws = NULL,
  cv_x_max = 150,
  pcov_threshold = 20,
  is_log_x = TRUE
)

Arguments

samples

Data frame of test samples.

bayes_fit

Output of fit_bayes_single().

curve_idx

Integer. Which curve (1-based Stan index).

response_variable

Character.

is_log_response

Logical.

n_draws

Integer or NULL.

cv_x_max

Numeric. Default 150.

pcov_threshold

Numeric. Percent CV threshold for pcov_pass. Default 20.

is_log_x

Logical. Default TRUE.

Value

Data frame with original sample columns plus prediction columns.