UniGASFor: Forecast with univariate GAS models
In GAS: Generalized Autoregressive Score Models

Description Usage Arguments Value Author(s) References Examples

Forecast with univariate GAS models. The one-step ahead prediction of the conditional density is available in closed form. The multi-step ahead prediction is performed by simulation as detailed in Blasques et al. (2016).

1 2	UniGASFor(uGASFit, H = NULL, Roll = FALSE, out = NULL, B = 10000, Bands = c(0.1, 0.15, 0.85, 0.9), ReturnDraws = FALSE)

uGASFit

An object of the class uGASFit created using the function UniGASFit.

`H`	`numeric` Forecast horizon. Ignored if `Roll = TRUE`.
`Roll`	`logical` Forecast should be made using a rolling procedure ? Note that, if `Roll = TRUE`, then `out` has to be specified.
`out`	`numeric` Vector of out-of-sample observation for rolling forecast.
`B`	`numeric` Number of draws from the H-step ahead distribution if `Roll = FALSE`.
`Bands`	`numeric` Vector of probabilities representing the confidence band levels for multi-step ahead parameters forecasts. Only if `Roll = FALSE`.
`ReturnDraws`	`logical` Return the draws from the multi-step ahead predictive distribution when `Roll = FALSE` ?

An object of the class uGASFor.

Leopoldo Catania

Blasques F, Koopman SJ, Lasak K, and Lucas, A (2016). "In-sample Confidence Bands and Out-of-Sample Forecast Bands for Time-Varying Parameters in Observation-Driven Models." International Journal of Forecasting, 32(3), 875-887. doi: 10.1016/j.ijforecast.2016.04.002.

# Specify an univariate GAS model with Student-t
# conditional distribution and time-varying location, scale and shape parameter

# Inflation Forecast

set.seed(123)

data("cpichg")

GASSpec = UniGASSpec(Dist = "std", ScalingType = "Identity",
                     GASPar = list(location = TRUE, scale = TRUE, shape = FALSE))

# Perform H-step ahead forecast with confidence bands

Fit = UniGASFit(GASSpec, cpichg)
Forecast = UniGASFor(Fit, H = 12)

Forecast

# Perform 1-Step ahead rolling forecast

InsampleData = cpichg[1:250]
OutSampleData = cpichg[251:276]

Fit = UniGASFit(GASSpec, InsampleData)

Forecast = UniGASFor(Fit, Roll = TRUE, out = OutSampleData)

Forecast

Attaching package: 'GAS'

The following object is masked from 'package:stats':

    residuals


------------------------------------------
-        Univariate GAS Forecast         -
------------------------------------------

Model Specification
Conditional distribution:  std
Score scaling type:  Identity
Horizon:  12
Rolling forecast:  FALSE
------------------------------------------
Parameters forecast:
      location     scale    shape
T+1 0.10130293 0.1523661 6.526135
T+2 0.09499435 0.1736636 6.526135
T+3 0.09381908 0.2151041 6.526135
T+4 0.09255762 0.2576731 6.526135
T+5 0.08746747 0.3019555 6.526135

....................
       location     scale    shape
T+8  0.08345437 0.4219074 6.526135
T+9  0.07792038 0.4574555 6.526135
T+10 0.07382608 0.4899598 6.526135
T+11 0.07558255 0.5198954 6.526135
T+12 0.07507434 0.5465449 6.526135

------------------------------------------
-        Univariate GAS Forecast         -
------------------------------------------

Model Specification
Conditional distribution:  std
Score scaling type:  Identity
Horizon:  26
Rolling forecast:  TRUE
------------------------------------------
Parameters forecast:
                     location     scale    shape CPIAUCSL
2009-09-30 22:00:00 0.8710313 0.4014086 6.374371  0.77988
2009-12-31 23:00:00 0.8407661 0.2959634 6.374371  0.15838
2010-03-31 22:00:00 0.6666247 0.3056644 6.374371 -0.03543
2010-06-30 22:00:00 0.5042359 0.3163431 6.374371  0.29272
2010-09-30 22:00:00 0.4572833 0.2470877 6.374371  0.80662

....................
                       location     scale    shape CPIAUCSL
2014-12-31 23:00:00  0.04129509 0.1138605 6.374371 -0.72630
2015-03-31 22:00:00 -0.25352898 0.1942819 6.374371  0.60266
2015-06-30 22:00:00  0.05156038 0.2760946 6.374371  0.34143
2015-09-30 22:00:00  0.18005571 0.2274846 6.374371  0.19128
2015-12-31 23:00:00  0.21543112 0.1789733 6.374371 -0.07815