Declare the Output Metric Schema for a Model

Returns a named list describing the metrics that fit_roi produces for this model type. Used by the generic combiner (Phase 2) to build the correct number of output maps without model-specific combine_* functions.

Usage

output_schema(model)

# S3 method for class 'manova_model'
output_schema(model)

# S3 method for class 'mvpa_model'
output_schema(model)

# S3 method for class 'item_model'
output_schema(model)

# S3 method for class 'rsa_model'
output_schema(model)

# S3 method for class 'vector_rsa_model'
output_schema(model)

# S3 method for class 'feature_rsa_model'
output_schema(model)

# S3 method for class 'naive_xdec_model'
output_schema(model)

# S3 method for class 'era_partition_model'
output_schema(model)

# S3 method for class 'repmap_model'
output_schema(model)

# S3 method for class 'repmed_model'
output_schema(model)

# S3 method for class 'repnet_model'
output_schema(model)

# S3 method for class 'subspace_alignment_model'
output_schema(model)

Arguments

model

A model specification object.

Value

A named list where names are metric names and values describe the type: "scalar" (one value per ROI) or "vector[N]" (N values per ROI). Returns NULL for models using the legacy combiner path.

Examples

# \donttest{
  # Default returns NULL (legacy path)
  ds <- gen_sample_dataset(c(4,4,4), 20)
  mdl <- load_model("sda_notune")
  mspec <- mvpa_model(mdl, ds$dataset, ds$design, "classification")
  output_schema(mspec)  # NULL
#> $Accuracy
#> [1] "scalar"
#> 
#> $AUC
#> [1] "scalar"
#> 
# }