Extract Features from FDA Significant Intervals — extract_fda

Converts significant FDA intervals into ordinary subject-level predictors for downstream modeling.

Usage

extract_fda_features(
  analysis,
  alpha = NULL,
  value_cols = NULL,
  prefix = "fda",
  statistics = "mean"
)

analysis: A windcut_fda_analysis object returned by run_fda_analysis().
alpha: Significance level used to select intervals. If NULL, the first alpha value stored in analysis is used.
value_cols: Optional subset of weather variables to extract.
prefix: Prefix used in feature names.
statistics: Interval statistics to compute. Use function names such as "mean", "sd", and "IQR", a named list of custom functions, or a named list by variable. When using a variable-level list, .default can supply statistics for variables not named explicitly.

A wide data frame with one row per subject and one column per extracted FDA feature.