- Added
generate()
andIRF()
methods for VAR models. - Added
IRF()
method for ARIMA models. - Added
VECM()
andVARIMA()
models.
- Documentation improvements.
Small patch to resolve issues in C++ R headers.
- Documentation improvements.
Small patch to resolve CRAN check issues.
- Documentation improvements.
- Fixed
generate(<ARIMA>)
method for some variable names. - Fixed df in
generate(<TSLM>)
.
- Documentation improvements.
- Added
approx_normal
argument toforecast(<TSLM>)
. This allows you to optionally return forecasts from the more appropriate Student's T distribution instead of approximating to a Normal distribution. The default behaviour remains the same, which is to provide approximate Normal distribution forecasts which are nicer to work with in model combination and reconciliation (#343). ETS()
will now ignore the smoothing parameter's range when specific parameter value is given (#317).- Modified initial parameter values for
ETS()
when bounds = "admissible". - Updated RW forecasts to use an unbiased estimate of sigma2 (#368).
- Fixed issue with characteristic equation test for admissibility of ETS parameters (#341).
- Fixed ARIMA selecting differences that don't satisfy the
order_constraint
(#360). - Fixed issue with forecasting ARIMA models with intercept and exogenous regressors.
- Fixed issue with VAR models not storing lagged regressor data for forecasting.
Small release to resolve check issues with the development and patched versions
of R. The release includes some minor improvements to the output consistency of
initial states in ETS()
models, the passing of arguments in ARIMA()
models,
and handling of missing values in NNETAR()
.
- Display of ETS initial states now use a
state[t]
notation to describe the state's position in time (#329, #261). - Allowed specifying
method
argument inARIMA()
(#330). - Improved handling of missing values in
NNETAR()
(#327).
- Fixed error with forecasting and simulating from
NNETAR()
estimated using a short series (#326). - Fixed
AR()
fitted values not being re-scaled to match original data (#318).
The release of fabletools v0.3.0 introduced general support for computing h-step
ahead fitted values, using the hfitted(<mdl>, h = ???)
function. This release
adds model-specific hfitted()
support to ARIMA and ETS models for improved
performance and accuracy.
This release adds improved support for refitting models, largely in thanks to contributions by @Tim-TU.
It is also now possible to specify an arbitrary model selection criterion
function for automatic ARIMA()
model selection.
- Added
refit()
method for NNETAR, MEAN, RW, SNAIVE, and NAIVE models (#287, #289, #321. @Tim-TU). - Added
hfitted()
method for ETS and ARIMA, this allows fast estimation of h-step ahead fitted values. - Added
generate()
method for AR, theforecast()
method now supports bootstrap forecasting via this new method.
- Added the
selection_metric
argument toARIMA()
, which allows more control over the measure used to select the best model. By default this function will extract the information criteria specified by theic
argument. - Added
trace
argument for tracing the selection procedure used inARIMA()
- Fixed unnecessary warning when forecasting short horizons using
NNETAR()
. - Fixed
generate()
method for NNETAR models when data isn't scaled (#302). - Fixed
refit.ARIMA()
re-selecting constant instead of using the provided model's constant usage. - Fixed use of exogenous regressors in
AR()
models.
This release coincides with v0.2.0 of the fabletools package, which contains
some substantial changes to the output of forecast()
methods.
These changes to fabletools emphasise the distribution in the fable
object. The most noticeable is a change in column names of the fable, with the
distribution now stored in the column matching the response variable, and the
forecast mean now stored in the .mean
column.
For a complete summary of these changes, refer to the fabletools v0.2.0 release
news: https://fabletools.tidyverts.org/news/index.html
- Added the
THETA()
method.
- Forecasts distributions are now provided by the distributional package. They
are now more space efficient and allows calculation of distributional
statistics including the
mean()
,median()
,variance()
,quantile()
,cdf()
, anddensity()
. - The uncertainty of the drift parameter in random walk models (
RW()
,NAIVE()
andSNAIVE()
) is now included in data generated withgenerate()
. - Added Syntetos-Boylan and Shale-Boylan-Johnston variants of
CROSTON()
method. - Performance improvements.
- Fixed issue with approximation being used when refitting ARIMA models and when a specific model is requested.
- Fixed
glance()
forTSLM()
models when the data contains missing values. - Fixed typo in
glance()
output ofETS()
models.
- The sample path means are now used instead of analytical means when forecasts are produced from sample paths.
- Added autoregressive modelling with
AR()
. - Better handling of rank deficiency in
ARIMA()
. - Added
generate.ARIMA()
method. - Added bootstrap forecast paths for
ARIMA()
models. ARIMA()
specials now allow specifying fixed coefficients via thefixed
argument.- Documentation improvements.
- Added
CROSTON()
for Croston's method of intermittent demand forecasting. - Documentation improvements
- Fixed NNETAR & VAR handling of missing values (#215).
- Fix ETS forecasting with forecast horizons less than the seasonal period (#219).
- Fixed season() special for non-seasonally based time indices (#220)
- Fix issue with simulation forecasting from damped ETS models.
- Added interpolation method for
MEAN()
model (#203). - Added rolling mean option for
MEAN()
model (#204).
- Corrected forecast standard error for drift models.
- First release.
- Support for 9 models and relevant methods:
ARIMA
,ETS
,TSLM
,MEAN
,RW
,NAIVE
,SNAIVE
,NNETAR
,VAR
.