posterior_predict: Outcome prediction from posterior distribution of parameters

Description

Compute outcome predictions using posterior samples. Exposure data for prediction can be either original data used for model fit or new data.

Usage

posterior_predict(object, ...)
posterior_epred(object, ...)
posterior_linpred(object, transform = FALSE, ...)
# S3 method for stanemax
posterior_predict(
  object,
  newdata = NULL,
  returnType = "matrix",
  newDataType = "raw",
  ...
)
# S3 method for stanemaxbin
posterior_predict(
  object,
  newdata = NULL,
  returnType = "matrix",
  newDataType = "raw",
  ...
)
# S3 method for stanemax
posterior_epred(object, newdata = NULL, newDataType = "raw", ...)
# S3 method for stanemaxbin
posterior_epred(object, newdata = NULL, newDataType = "raw", ...)
# S3 method for stanemax
posterior_linpred(
  object,
  transform = FALSE,
  newdata = NULL,
  newDataType = "raw",
  ...
)
# S3 method for stanemaxbin
posterior_linpred(
  object,
  transform = FALSE,
  newdata = NULL,
  newDataType = "raw",
  ...
)
posterior_predict_quantile(
  object,
  newdata = NULL,
  ci = 0.9,
  pi = 0.9,
  newDataType = c("raw", "modelframe")
)

Value

An object that contain predicted response with posterior distribution of parameters. The default is a matrix containing predicted response for stan_emax() and .epred for stan_emax_binary(). Each row of the matrix is a vector of predictions generated using a single draw of the model parameters from the posterior distribution.

If either dataframe or tibble is specified, the function returns a data frame or tibble object in a long format - each row is a prediction generated using a single draw of the model parameters and a corresponding exposure.

Several types of predictions are generated with this function.

For continuous endpoint model (stan_emax()),

.linpred & .epred: prediction without considering residual variability and is intended to provide credible interval of "mean" response.
.prediction: include residual variability in its calculation, therefore the range represents prediction interval of observed response.
respHat: replaced by .linpred and .epred
response: replaced by .prediction

For binary endpoint model (stan_emax_binary()),

.linpred: predicted probability on logit scale
.epred: predicted probability on probability scale
.prediction: predicted event (1) or non-event (0)

The return object also contains exposure and parameter values used for calculation.

With posterior_predict_quantile() function, you can obtain quantiles of respHat and response as specified by ci and pi.

Arguments

object: A stanemax or stanemaxbin object
...: Additional arguments passed to methods. Arguments that can be passed via the dots include ndraws, for compatibility with functions in the tidybayes package
transform: Should the linear predictor be transformed to response scale?
newdata: An optional data frame that contains columns needed for model to run (exposure and covariates). If the model does not have any covariate, this can be a numeric vector corresponding to the exposure metric.
returnType: An optional string specifying the type of return object (one of "matrix", "dataframe", or "tibble")
newDataType: An optional string specifying the type of newdata input, whether in the format of an original data frame ("raw", the default) or a processed model frame ("modelframe"). Mostly used for internal purposes and users can usually leave at default.
ci: Credible interval of the response without residual variability.
pi: Prediction interval of the response with residual variability.

Details

Run vignette("emaxmodel", package = "rstanemax") to see how you can use the posterior prediction for plotting estimated Emax curve.