Iterates over all (or selected) properties for family_id and loads
each model into the in-memory cache. Call this once at Shiny app startup to
eliminate per-prediction disk I/O.
Usage
preload_ossl_models(
family_id,
model_dir = getOption("autoSpectra.model_dir", "models"),
properties = NULL,
verbose = TRUE
)
Arguments
- family_id
Family identifier ("OSSL_VisNIR" or "OSSL_MIR")
- model_dir
Root directory where trained models are stored
- properties
Character vector of properties to pre-load.
NULL (default) loads all properties defined for the family.
- verbose
Print progress messages (default TRUE)
Value
Invisible character vector of successfully loaded property names
