larchEstfun

Two-step estimation of a log-variance model: OLS in step 1, bias correction w/residuals in step 2 (see the code for details). The function <code>larchEstfun()</code> is not intended for the average user, but is called by <code>larch</code> and <code>gets.larch</code>.

time series

Statistical Models

Econometrics

Financial Econometrics

Climate Econometrics

Automated General-to-Specific (GETS) modelling of the mean and variance of a regression, and indicator saturation methods for detecting and testing for structural breaks in the mean, see Pretis, Reade and Sucarrat (2018) <doi:10.18637/jss.v086.i03> for an overview of the package. In advanced use, the estimator and diagnostics tests can be fully user-specified, see Sucarrat (2021) <doi:10.32614/RJ-2021-024>.

Genaro Sucarrat

gets

General-to-Specific (GETS) Modelling and Indicator Saturation
Methods

larchEstfun function

<dl> <dt>loge2</dt>
<dd>numeric vector, the log of the squared errors 'e' (adjusted for zeros on e, if any)</dd> <dt>x</dt>
<dd>numeric matrix, the regressors</dd> <dt>e</dt>
<dd>numeric vector, the errors</dd> <dt>vcov.type</dt>
<dd><code>character</code> vector, "robust" (default) or "hac". If "robust", then the White (1980) heteroscedasticity-robust variance-covariance matrix is used for inference. If "hac", then the Newey and West (1987) heteroscedasticity and autocorrelation-robust matrix is used</dd> <dt>tol</dt>
<dd>numeric value. The tolerance for detecting linear dependencies in the columns of the regressors in the first step estimation by OLS, see <code>ols</code>. Only used if <code>LAPACK</code> is <code>FALSE</code></dd></dl>

Arguments

Genaro Sucarrat, <a href="http://www.sucarrat.net/">http://www.sucarrat.net/</a>

Author

Estimation of a log-variance model — larchEstfun

<dl>

 <dt>loge2</dt>
<dd>numeric vector, the log of the squared errors 'e' (adjusted for zeros on e, if any)</dd>

 <dt>x</dt>
<dd>numeric matrix, the regressors</dd>

 <dt>e</dt>
<dd>numeric vector, the errors</dd>

 <dt>vcov.type</dt>
<dd><code>character</code> vector, "robust" (default) or "hac". If "robust", then the White (1980) heteroscedasticity-robust variance-covariance matrix is used for inference. If "hac", then the Newey and West (1987) heteroscedasticity and autocorrelation-robust matrix is used</dd>

 <dt>tol</dt>
<dd>numeric value. The tolerance for detecting linear dependencies in the columns of the regressors in the first step estimation by OLS, see <code>ols</code>. Only used if <code>LAPACK</code> is <code>FALSE</code></dd>

</dl>

Genaro Sucarrat, <a href='http://www.sucarrat.net/'>http://www.sucarrat.net/</a>

larchEstfun: Estimation of a log-variance model

Description

Usage

Value

Arguments

Author

Details

References

See Also

Examples