narx

Nonlinear autoregressive exogenous (NARX) model for system identification.

Classes

NARX(*[, feat_ids, delay_ids, output_ids, ...])

The Nonlinear Autoregressive eXogenous (NARX) model class.

Functions

fd2tp(feat_ids, delay_ids)	Convert feat_ids and delay_ids to time_shift_ids and poly_ids.
gen_poly_features(X, ids[, skip_indices])	Generator to make polynomial features.
gen_time_shift_features(X, ids[, skip_indices])	Generator to make time shift features.
make_narx(X, y, n_terms_to_select[, ...])	Find time_shift_ids, poly_ids, output_ids for a NARX model.
make_poly_features(X, ids)	Make polynomial features.
make_poly_ids([n_features, degree, ...])	Generate ids for polynomial features.
make_time_shift_features(X, ids, **kwargs)	Make time shift features.
make_time_shift_ids([n_features, max_delay, ...])	Generate ids for time shift features.
print_narx(narx[, term_space, coef_space, ...])	Print a NARX model as a Table which contains y idx, Term, and Coef.
tp2fd(time_shift_ids, poly_ids)	Convert time_shift_ids and poly_ids to feat_ids and delay_ids.