Skip to contents

Run `multifer` inference for a fitted [aligned_rrr()] model. The adapter uses multifer's opaque predictive geometry so the model can keep its native multiblock predictor/response data shape.

Usage

infer_aligned_rrr(
  x,
  data = NULL,
  targets = "default",
  B = 1000L,
  R = 500L,
  alpha = 0.05,
  seed = NULL,
  strict = TRUE,
  ...
)

Arguments

x

A fitted `aligned_rrr` object.

data

Optional original data list with elements `X` and `Y`. When `NULL`, the data stored in `x$fit_spec$refit$data` are used.

targets

Inference targets passed to [multifer::infer()]. Defaults to all targets declared by the adapter: component significance plus loading, score, and subspace stability.

B

Number of Monte Carlo draws per component-significance rung.

R

Number of bootstrap replicates for stability targets.

alpha

Significance level passed to [multifer::infer()].

seed

Optional random seed.

strict

Logical; passed to [multifer::infer()].

...

Additional arguments passed to [multifer::infer()].

Value

A `multifer::infer_result` object, or a `multifer` bundle when `return_bundle = TRUE` is passed through `...`.