simulate.GNARfit

Simulates from a GNARfit object, either creating a new series or future observations of the original series based upon the fitted GNAR model.

Simulation of, and fitting models for, Generalised Network Autoregressive (GNAR) time series models which take account of network structure, potentially with exogenous variables. Such models are described in Knight et al. (2020) <doi:10.18637/jss.v096.i05> and Nason and Wei (2021) <doi:10.1111/rssa.12875>. Diagnostic tools for GNAR(X) models can be found in Nason et al. (2023) <doi:10.48550/arXiv.2312.00530>.

Matt Nunes

GNAR

Methods for Fitting Network Time Series Models

Kathryn Leeming

Guy Nason

Marina Knight

James Wei

Daniel Salnikov

Mario Cortina Borja

simulate.GNARfit function

<dl> <dt>object</dt>
<dd>the output of a GNARfit call</dd> <dt>nsim</dt>
<dd>the time length of the simulations</dd> <dt>seed</dt>
<dd>either <code>NULL</code>, or a value to set the seed to</dd> <dt>future</dt>
<dd>whether the simulations follow on from the original time series (<code>TRUE</code>), or if <code>FALSE</code> the simulations are a new series.</dd> <dt>set.noise</dt>
<dd>a value to set the standard deviation of the noise to, or if <code>NULL</code>, the estimated standard deviation from the input series will be used.</dd> <dt>allcoefs</dt>
<dd>if <code>TRUE</code>, all fitted coefficients will be used, if <code>FALSE</code> only the significant (p-val &lt; 0.05) coefficients will be used.</dd> <dt>...</dt>
<dd>additional arguments, unused here.</dd></dl>

Arguments

Function to simulate from a GNARfit object — simulate.GNARfit

<dl>

 <dt>object</dt>
<dd>the output of a GNARfit call</dd>

 <dt>nsim</dt>
<dd>the time length of the simulations</dd>

 <dt>seed</dt>
<dd>either <code>NULL</code>, or a value to set the seed to</dd>

 <dt>future</dt>
<dd>whether the simulations follow on from the original time series (<code>TRUE</code>), or if <code>FALSE</code> the simulations are a new series.</dd>

 <dt>set.noise</dt>
<dd>a value to set the standard deviation of the noise to, or if <code>NULL</code>, the estimated standard deviation from the input series will be used.</dd>

 <dt>allcoefs</dt>
<dd>if <code>TRUE</code>, all fitted coefficients will be used, if <code>FALSE</code> only the significant (p-val &lt; 0.05) coefficients will be used.</dd>

 <dt>...</dt>
<dd>additional arguments, unused here.</dd>

</dl>

simulate.GNARfit: Function to simulate from a GNARfit object

Description

Usage

Value

Arguments

Details

Examples