Package index • ffc

Core Functions

Main functions for functional forecasting

ffc_gam(): Fit a functional time series model using dynamic functional coefficients
fts(): Define functions with dynamic coefficients in ffc formulae

Model Methods

Methods for working with fitted models

predict(<ffc_gam>): Predict from a fitted ffc_gam model
forecast(<ffc_gam>): Forecasting ffc_gam models
forecast(<fts_ts>): Forecasting functional basis coefficients (Internal)
model.frame(<ffc_gam>): Extract model.frame from a fitted ffc_gam object

Coefficient Extraction

Extract and work with time-varying coefficients

fts_coefs(): Extract time-varying basis coefficients
autoplot(<fts_ts>): Plot time series of fts basis coefficients
print(<fts_ts>): Print an fts_ts tibble

Forecasting Models

Specialized forecasting models for time-varying coefficients

ARDF(): Fit an autoregressive dynamic factor model
GPDF(): Fit a GP dynamic factor model
VARDF(): Fit a Vector autoregressive dynamic factor model

Integration with fabletools

Convert ffc objects to work with the fabletools ecosystem

as_fable(<ffc_gam>): Convert ffc_gam forecasts to fable object
reexports s te gam bam nb negbin betar cnorm ocat scat ziP multinom tw ldTweedie rTweedie twlss autoplot forecast generate: Objects exported from other packages

Data

Example datasets included with the package

growth_data: Berkeley growth study data
qld_mortality: Queensland mortality data
elnino_sst: El Niño Sea Surface Temperature Data

Re-exported Functions

Functions re-exported from mgcv and other packages for convenience

reexports s te gam bam nb negbin betar cnorm ocat scat ziP multinom tw ldTweedie rTweedie twlss autoplot forecast generate: Objects exported from other packages