predict.gctsc: One-Step-Ahead Predictive Distribution for Copula Count Time...

In gctsc: Gaussian and Student-t Copula Models for Count Time Series

One-Step-Ahead Predictive Distribution for Copula Count Time Series Models

Description

Computes the one-step-ahead predictive distribution for a fitted Gaussian or Student–t copula count time series model.

The predictive probability mass function is evaluated using the estimation method stored in the fitted object (TMET or GHK). Summary statistics of the predictive distribution are returned, and optional scoring rules are computed if the observed value is supplied.

Usage

## S3 method for class 'gctsc'
predict(object, y_obs = NULL, X_test = NULL, ...)

Arguments

object	A fitted model object of class "gctsc", as returned by gctsc.
y_obs	Optional non-negative integer giving the observed value at the prediction time. If supplied, the Continuous Ranked Probability Score (CRPS) and Logarithmic Score (LOGS) are computed.
X_test	Covariate values at the prediction time point. Can be provided as: a named numeric vector whose names match the covariates used during model fitting, or a one-row data.frame or matrix with matching column names. If the fitted model is intercept-only, X_test may be omitted, in which case the intercept is set to 1 automatically. An error is raised if more than one row is supplied.
...	Ignored. Included for S3 method compatibility.

Details

The predictive distribution integrates over the latent copula dependence structure specified in the fitted model. For Gaussian copulas, multivariate normal rectangle probabilities are evaluated; for Student–t copulas, multivariate t rectangle probabilities are used.

Value

A list containing:

• mean: Predictive mean.

• median: Predictive median.

• mode: Predictive mode.

• variance: Predictive variance.

• p_y: Predictive probability mass function over 0:y_max.

• lower, upper: Bounds of the 95\

• CRPS: Continuous Ranked Probability Score (if y_obs is provided).

• LOGS: Logarithmic Score (if y_obs is provided).

References

Nguyen, Q. N., & De Oliveira, V. (2026). Likelihood Inference in Gaussian Copula Models for Count Time Series via Minimax Exponential Tilting Journal of Computational Statistics and Data Analysis.

Nguyen, Q. N., & De Oliveira, V. (2026). Scalable Likelihood Inference for Student–t Copula Count Time Series. Manuscript in preparation.

See Also

gctsc, arma.cormat

Examples

# Simulate Poisson AR(1) data
set.seed(1)
y_sim <- sim_poisson(mu = 10, tau = 0.2,
                     arma_order = c(1, 0),
                     nsim = 1000,
                     family = "gaussian")$y

# Fit Gaussian copula model
fit <- gctsc(
  y ~ 1,
  data = data.frame(y = y_sim),
  marginal = poisson.marg(lambda.lower = 0),
  cormat = arma.cormat(p = 1, q = 0),
  method = "CE",
  family = "gaussian",
  options = gctsc.opts(M = 1000, seed = 42)
)

# One-step-ahead prediction
predict(fit)

gctsc documentation built on March 20, 2026, 9:11 a.m.