If reticulate and numpy are available, computes auxiliary
dynamics of the valence series (gradient and cumulative trajectory) in
Python and attaches them to the score's enrichment slot. The deterministic
core features, arc, and mapping are left untouched — enrichment is additive
and never feeds the reproducibility snapshot.
Usage
syn_py_features(score, call = rlang::caller_env())Arguments
- score
A
prosody_scorefromsyn_score().- call
The calling environment, for error reporting. Experts only.
Value
The score with score$enrichment$py_features added (a tibble of
sentence_id, valence_gradient, valence_cumulative). Requires
reticulate + Python numpy.
See also
Other assist functions:
syn_export_video(),
syn_interpret(),
syn_llm_condense(),
syn_suggest()
Examples
if (FALSE) { # \dontrun{
# Not run: needs the reticulate package and a Python installation with
# numpy, which are not available on every system.
score <- syn_score(syn_read_text(syn_example_text("en"), lang = "en"))
syn_py_features(score)$enrichment$py_features
} # }