Skip to contents

Fit Reference Model for Projective Prediction

fit_reference.Rd

Using a regularised horseshoe prior, fits a full reference model of selected "gold standard" PDD diagnosis using specified predictor set.

Usage

fit_reference(
  d,
  gs = "Lvl.II (1)",
  y = "PDD",
  X = c("moca_7", "moca_cloc", "vf_k", "moca_5words", "moca_cube", "moca_abs",
    "moca_anim", "mmse_7", "cloc", "vf_s", "mmse_3words", "mmse_pent"),
  expect_nonzero = 5,
  ...
)

Arguments

d

Data such as those calculated by diagnose_pdd_sample.

gs

Name the "gold standard" variable has in d$type.

y

Name of the PDD diagnosis variable in d

X

Character vector containing all predictor variables from d.

expect_nonzero

Number of variables expected to be truly predictive.

...

Other parameters for brms::brm().

Value

List with three components:

model

brmsfit with the fitted model

data

Tibble with raw input data

scaling

Tibble with predictors' mean and SDs

Details

...

See also